The impression that science is going through a new phase of reorganization and retrenchment is widespread, and growing (Brooks, 1993; Etzkowitz & Webster, 1995; Ezrahi, 1990; Forman, 1997; Guston & Keniston, 1994; Wade, 1984; Ziman, 1994). In fact, we are witnessing the transition to a new regime, characterized by an expanded scale of scientific activities as a result of sophistication and collectivization, the rise of customer-contractor relationships, changing professional roles and career paths that blur the distinction between academia and the "outside" world, shifting boundaries between pure and applied science and between science and technology, and so on. In addition, science is experiencing the effects of general cultural, political, and economic changes. For instance, the political and ideological role of science is eroding, there is a strong desire to justify the large sums of public money that pour into modern science, the relationship among business, the military, the government, and the university is undergoing intense reevaluation, etc.

Several stories concerning these developments are circulating. Some attribute them to the fact that science was growing too fast in relation to its underlying support system (Ziman, 1994). Others argue that there are inherently diminishing returns to research (Rescher, 1989). Yet others believe the reduced interest in military research and development combined with increased international competition are at fault (Slaughter & Rhoades, 1996). Through a complex mix of internal and external causes and effects, these developments are having a profound impact on the funding of science, the organizational structure of science, scientific careers, the multinational character of science, and so on.

In such a climate characterized by scrutiny and control of science, it is not too surprising to encounter appeals to an "economics of science" (as contrasted with an economics of technological change) that will serve to structure the inchoate impressions of the various constituencies involved, as well as to provide a basis for reasoned debate and guidance for public policy. The choice is either to deny that the economic transformations we are experiencing are having any profound influence upon science and the academy, and live in an atmosphere of increasingly harsh recriminations, or else to mobilize some broad-based economic expertise in order to analyze and respond to the changes. A deeper understanding of measurement and evaluation of scientific research might aid policymakers and funding agencies in making their decisions. In several areas, the discussions of the economic value(s) in and of science can provide an alternative perspective to the analyses taking place in science studies. For example, it can help answer the questions raised during discussions about conduct and evaluation of science policy. What kinds of considerations should go into funding decisions? In addition, the economics of science has an important role to play in redesigning institutions of higher education and research for the 21st century. Universities and industry are assuming tasks that were formerly largely the province of the other. A new social contract is being drawn up between the university and the larger society (Guston & Keniston, 1994), in which public funding for the university is made contingent upon a more direct contribution to the economy. Has economic development become a new rationale of the university in addition to teaching and research? If so, how does one measure and evaluate this function, and what impact will it have upon the structure of science?

The goal of this paper is to provide an overview of the role of the economic value(s) in and of science. The next section provides a historical perspective in terms of distinct regimes of science organization and funding in the United States in the 20^th century. The third section outlines several possible philosophical positions of economics of science.

Based upon a cross-section of recent work in the history of science and science policy, we would like to propose that there have existed three very distinct regimes of science organization and funding in the United States in the 20^th century, and that each regime has borne a special relationship to the contemporary organization of science in other countries. While they may overlap in certain particulars, we will place the temporal divisions of these regimes at the following boundaries: early 20^th century-1940; between World War II and the Cold War; and (roughly) 1980-present. For convenience, we will call them the (1) proto-industrial regime, (2) the Cold War regime, and (3) the globalized privatization regime.

The first, proto-industrial regime of science funding and organization, dates from an era when American science was widely regarded as inferior in many respects to European science, and when a few American universities were entering their initial phase as incubators of scientific research. In this era, most colleges and universities existed almost exclusively to perform the service of education and the propagation of the liberal arts, based upon Scots and British models of liberal education. As our label intimates, most scientific research and development in this period was to be found in a few large American corporations. The reasons that some large corporations such as General Electric, DuPont, American Telephone and Telegraph, and Eastman Kodak fostered in-house research capacity had much more to do with the need for routine testing capacities and with the fin de siècle merger wave and American anti-trust and patent policies than with any belief in the necessity of innovation or the commercial value of science, as is now widely acknowledged in the historical literature. America lagged behind European practice only by a few decades, since the modern literature dates the inception of corporate research labs for the purpose of product innovation in Europe to roughly the 1880s (Fox & Guagnini, 1999, pp. 215, 251). In the public sphere, the US Federal government role in supporting research was comparatively small, and consisted primarily of promotion of agricultural research through a network of agricultural extension stations, or else subsidizing specialized research in government-run labs tied to motives of nation-building. Examples of the latter would include the Coast Survey (which employed the earliest author in this anthology, Charles Sanders Peirce), the US Geological Survey, the Bureau of Chemistry of the Agriculture Department, and the National Bureau of Standards. In many respects, American policy reformers sought (with indifferent success) to mimic science policies first innovated in Germany, especially in the promotion of state-funded higher education combined with state-run research institutes, and with good reason, since German science was thought to be the best in the world (Lenoir, 1998). What the German system had innovated under the aegis of the Humboldtian reforms was a more closer integration of (graduate) teaching and research, which extended to the institution of a laboratory-based pedagogy (Fox & Guagnini, 1998/9). American students seeking advanced academic training were therefore urged to spend time in German universities, in the absence of suitable American infrastructure. The professionalization of academic disciplines was then in its earliest stages; and in most fields, a career consisting primarily of research was simply not a viable option. Unless one worked for a Federal or a corporate lab, the life of an American scientist was a hard one. Only toward the very end of the period did a handful of private foundations (such as Rockefeller and Carnegie; see Kohler, 1991) begin to innovate new forms of scientific patronage, aimed at building up a few selected universities as research institutions, and revising the previous construction of the research grant as a temporary dole parceled out as charity to poverty-stricken academics. Thus both corporate science and the nascent structure of academic careers in America were almost entirely defined by the captains of industry and their managers.

As noted by Reingold (1991), Americans had great trouble coming to terms with the nascent idea of public funding for a scientific elite. By the 1920s, there arose a substantial cultural trend that regarded industrial concentration and technological advancement as two sides of the same coin, a dynamic resulting in technological displacement, and in the 1930s, even widespread unemployment. Moreover, Continental Europeans (with the British emphatically excluded) tended to treat their scientists as having a status patterned upon their previous aristocracies, and thus took for granted their role within the state; but this option was foreclosed in the American context. Suspicion of elites tended to shade over into skepticism over the very premise that there should exist a cadre of researchers who would do their thinking for the benefit of the larger populace. This was captured by some comments in the New York Times of 1885:

Hence, we should note that there is nothing particularly novel or radical about the American penchant for regarding the scientist as a rational self-interested agent; nor is the notion of a market-driven science especially innovative or path-breaking. Indeed, it constituted the intellectual underpinnings for the bulk of science support in America at the turn of the last century. The point we wish to stress is that the science supported under this regime was relatively modest and rarely attained world-class status. Innovative American scientists tended toward being autodidacts and loners (Peirce himself serving as an extreme example), even if they managed to ascend to a university position, and many did not. It took a dramatic change of regime to propel American science to the front ranks of world science.

Because this spotty situation barely qualified as a "system", it should perhaps come as no surprise that there was very little literature dating from this era which could qualify as propounding a self-conscious "economics of science"; with the exception of a short article by Peirce (1879).

It is now almost universally acknowledged that World War II stands as the watershed of American science, and that the system of science funding and management in the United States propelled it to world dominance for the second half of the century. This is not to say that the feat was accomplished consciously and intentionally; nor is it to ignore the fact that the Depression and Nazi-era disruption conveniently destroyed the hitherto-dominant German university system and drove a generation of stellar scientific talent to seek asylum in the US. For obvious reasons, here is not the place to rehearse the familiar narrative of how World War II became known as the "physicist’s war," and how the postwar politics of the bomb and the Cold War locked America into a military-dominated system of science funding. Rather, we merely wish to indicate the ways in which the various components of the Cold War regime that were forged in the fires of WWII tended to fit together, and the ways in which this system was virtually unprecedented amongst the other economically developed nations, at least until the 1980s.

The centerpiece of this regime was the massive Federal government presence in science planning and funding. The dominance can be demonstrated by means of many quantitative measures, showing Federal expansion from the immediate postwar period to the mid-1960s, and then subsequent contraction, as illustrated in Figure 1 below.

But it was also a structural dominance, with the wartime practice (innovated at the Office of Scientific Research and Development) of research support in the format of government "contracts" being granted through universities and industrial firms on a "nonprofit" but fully reimbursed basis, strengthening both these institutions by using them to channel support to individual scientists, rather than simply hiring the scientists as civil service employees of a Federal laboratory system. This infusion of cash jump-started the cultivation of Big Science on a scale previously unimaginable, with massive instrumentation and hierarchical teams of interdisciplinary scientists and engineers, and patterned upon the successes of the MIT Radiation Laboratory and the Manhattan Project. Curiously enough, this was combined with a strong reinforcement of the centrality of the academic discipline as the arbiter of legitimacy of training and career success of the scientist. The vehicle for reconciling these seemingly contrary trends was the reconfigured postwar American university, where Federal science policy created a situation in which teaching was now openly avowed to be complementary to research (but not in those massive introductory courses staffed by graduate students); where "overhead costs" on contracts helped fund the operating expenses and the graduate population; while the GI bill and the creation of the temporary occupation known as the "research assistantship" forged a whole new career trajectory for novice scientists, and underwrote a massive expansion in graduate education. If some especially favored scientists still chafed under this shotgun marriage of teaching and research and disciplinary identity, a hybrid largely absent in most other developed economies, then a novel quasi-governmental entity known as the "think tank" – or more colloquially, a university campus without students – was created to cushion the irritation for the favored few.

It would be a mistake, however, to see American science policy as focused solely or even primarily upon universities. Federal and military science policies also fostered what has been called a "stealth industrial policy" in the US. It was an ideological imperative in America that the government not be seen as favoring certain industries in contradiction to the marketplace, a ruse that Cold War security considerations facilitated. Nevertheless, most Federal R&D funding was channeled through private corporations, even at the peak of university support, skewing the direction of technological exploration in selected industries. Private industry (broadly defined) has always been far and away the largest performing sector of R&D in the US, as revealed in Table 1.

Indeed, these figures understate the magnitude of Federal subsidy of corporate science, since a fair proportion of it occurred through tax rebates and third party payments. (It is an artifact of the Cold War regime that there exist no dependable consolidated R&D accounts for the entire Federal government, even down to the present. Things have, if anything, gotten worse with the decline of the Cold War regime, with the Department of Defense no longer providing detailed breakdowns of its own R&D spending by academic field after 1993.)

This industrial policy extended well beyond monetary grants and subsidies, however. In sharp contrast to the proto-industrial regime, the Cold War regime was characterized by a very weak legal structure of intellectual property protection, combined with a very active antitrust posture. The net consequence of this mode of science organization was that many of the scientific and technological breakthroughs achieved by corporate labs such as Bell Labs, Xerox Parc, RCA Sarnoff, Merck Rahway and IBM Yorktown were not adequately capitalized upon by their huge corporate sponsors, but were instead turned into downstream marketable commodities by small startup firms, themselves often formed by fugitives from those very same corporate labs. This "communal" approach to appropriation of the fruits of subsidized research was also encapsulated in the Department of Defense policy of a "second source rule" for suppliers of high-tech weaponry and devices, duly sweetened by cost-plus contracts. This rather cavalier attitude towards technology transfer from the laboratory to the marketplace was one of the prime hallmarks of the Cold War regime, one that could trace its provenance to the looming presence of the military in science funding.

The mutual reinforcement of this stealth industrial policy and the postwar ideology of the "freedom" of the scientist is another phenomenon, like the existence of a stealth industrial policy and the structure of intellectual property, that we believe has not yet been adequately explored in the science studies literature. The scientists most heavily embroiled in military funding had to submit to the classification and clearance procedures of the state; and in exchange, the state would promise not to micromanage their research agenda. Control, while not completely internalized, was certainly rendered unobtrusive; and the ability to appeal to freedom of expression was a critical component of the ideological rivalry of the period. Since a major feature of the Cold War regime was the maintenance of ongoing university ties of scientists doing government-sponsored work, they were exhorted also to publish in the "open" literature to meet their disciplinary obligations and bolster their credentials, if they exercised prudent discretion. (The existence of completely classified "scientific journals" stands as one of the more extreme anomalies of that era.) The "uses" of various discoveries therefore became more radically separated from their original elaborations (as well they might, given the rather imprecise concepts of intellectual property), especially in the formats in which they were disseminated. The economics of these disciplinary outlets were themselves often obscured through such indirect devices as page charges, submission fees, and wildly inflated library subscription rates. Crudely, it became possible for nominal academic scientific stature to be denominated in terms of public intangibles like disciplinary "credit" or "eminence," all the while the money was being allocated according to somewhat different criteria. Even though science was being closely managed by research officers, at first in the wartime Office for Scientific Research and Development (OSRD), and after the war in the Office of Naval Research, DARPA (Defense Advanced Research Projects Agency; or other permutations), the Atomic Energy Commission and elsewhere, the scientists eventually learned to come to terms with any residual sense that there might fester some conflict between their own freedom of inquiry and larger decisions to channel research in certain directions. If the exigencies of national security did not appear sufficiently compelling, the researcher could always take succor from the ethos of "pure science" within the academy. Indeed, this became the background to the public face of science policy enshrined in Vannevar Bush’s famous 1945 manifesto, Science—the Endless Frontier, namely, the "linear model" of "basic" science ® "applied" science ® "development" ® production. The alert reader will have detected that, until now, we have not acknowledged the existence of any hard and fast distinction between "basic vs. applied" science and "technology", for reasons that should now become clear. We regard the endless fascination in science studies (and, as we shall shortly observe, economics) with boundary maintenance between "basic" and "applied" science to be itself an artifact of the Cold War regime (Kline, 1995). This doctrine, so taken for granted within modern orthodox economics and much of orthodox science policy, was hardly even present in economic writings prior to World War II.

The Bush report has been analyzed repeatedly, and perhaps to excess, in the literature on the history of science policy. Our only concern here is to suggest that it played an important conceptual and ideological role in the Cold War regime, even if it did not end up serving as a blueprint for the actual structures of science funding and management that were eventually instituted in the US in this period. The idea that there was some "necessary but unproductive" form of scientific research that required state funding for its very existence, and that the economic growth of the nation would suffer in its absence, whereas applied R&D could be safely left to the corporate sector to organize, provided the ideal cover for the absence of accountability of military science planning in conjunction with the previously described stealth industrial policy that had precipitated out of the immediate postwar political process. Although often pitched at a rarefied level of abstraction seemingly free of any parochial considerations, it has only become somewhat clearer in retrospect that it was nonetheless a product of local conditions prevalent in the specific postwar regime in the US, and in fact bore little relevance for science policy in other countries in the same time frame.

Whether it be the cataclysmic downsizing of physics in the last two decades, or sweeping changes in the rules of the game for academic entrepreneurship, or the radical restructuring of research universities, everyone now realizes to a greater or lesser degree that the Cold War structure of science management is rapidly going the way of the whalebone corset and the phonograph record. Tectonic shifts of science funding between various sciences, as indicated below in Table 2, have been accompanied by drastic reorganization in the very structures of scientific funding and conduct.

From many different vantage points, it should now become apparent that the Cold War regime of science policy could not have persisted over the longer term. The convenient fiction of a clear separation between "pure" and "applied" science could not be long maintained. There were simply too many internal contradictions and repressed economic considerations in what had initially seemed a politically viable set of compromises. First, it was inevitable that the heavy subsidies provided by the Federal government to the universities and corporations would sooner or later have run into political and economic obstacles in the US political culture. Most Americans had never really relinquished their suspicions concerning coddled intellectual elites (Hollinger, 1995), and the trope of the danger of the self-interested and therefore untrustworthy scientific expert could easily be revived as a democratic plea for greater accountability in value for money. Furthermore, because the R&D budget of the Federal and state governments was so widely dispersed amongst numerous agencies and programs, and so devoid of coordination and effective interest-group mobilization (with the possible exception here of biomedical research), they made an inviting target if and when budgetary stringencies would prompt belt-tightening measures. A high-profile example of this growing vulnerability came with the cancellation of the Superconducting Super Collider in October 1993, an event that marks the dethronement of physics as the unchallenged champion of the Cold War regime (Sarewitz, 1996), even though Table 2 shows physics as a funding priority had eroded even earlier.

The supposed immunity of the quotidian prosecution of science from economic considerations was further compromised when universities themselves stopped being perceived as otherworldly ivory towers removed from politics, and were caught in the unseemly act of trampling all over each other in competition to get Federal research funds specifically earmarked to their campuses, lobbying representatives not on grounds of some abstract scientific "peer review," but rather for geographic or other political justifications (Brainard & Cordes, 1999). Another doomed holdover from the Cold War regime was the extreme concentration of research funds in a small number of elite institutions – Bush’s own MIT being the most favored recipient – a skewed distribution that could not persist once excluded universities realized that they could actively enter the political sweepstakes for Federal funds, and could effectively challenge previously hidden old boy networks. Indeed, "peer review" was one of the early casualties of the breakdown of the Cold War regime, since the insistence upon functioning internal standards of quality control could not begin to assuage external demands for responsibility, relevance, and accountability.

Third, the Cold War premise of "science policy in one nation" could not continue to be maintained in a world that was growing less bipolar, and more economically developed. Contacts between scientists across the Iron Curtain could be monitored in the name of military security; but as European and Asian firms regrouped and became economically significant, they conceived a desire to tap into the scientific and technological developments that had become such a prominent feature of postwar American prosperity. Since their own indigenous research infrastructures were so divergent from that found in the United States – frequently one set of institutions dedicated purely to instruction, another different set for state-sponsored research, and a third set charged with state planning of industrial research – they initially had to send their most promising students to the US to partake of the novel developments within the framework of the unfamiliar pedagogy. American universities were initially inclined to welcome the newcomers, but this had perverse unintended effects on American science policy. As the American research infrastructure grew so prodigiously up through the 1960s, it became apparent that the pool of indigenous candidates for scientific careers would not keep pace; and so foreign students were increasingly recruited into American academe and industrial research. Although wonderfully beneficial for American culture, this did tend to create problems for the prevalent political rationale for the government funding of higher education and its integration within the research system, which tended to be phrased in terms of investment in national human capital. It became increasingly difficult to justify the subsidized training of foreign students, many of whom would return to their home countries in order to staff the major economic and political competitors to the American system; and this does not even take into account the undercurrent of xenophobia undermining political support for government-subsidized education. As Figure 2 reveals, some disciplines began to be dominated by foreign nationals by the 1990s.

The net result was that nationalist pride and xenophobia were effectively undercut as rallying cries for American science policy from many different directions, and the emotional center of the national discourse about science congealed into a fear that convergence of other economies to US standards and practices (inevitable to some extent in any case) was undermining American economic "competitiveness."

The fourth countervailing tendency, materially the least gradual and therefore the most obvious in retrospect, was the utter collapse of the Soviet Union as the premier rival in the Cold War system. Because of its fragmented and partially classified nature, it had been previously impossible to gauge the extent of the importance of the national security imperative for the framework of science organization in America, at least until the fall of the Wall. Subsequently, many analysts have come to an increasingly nuanced appreciation of just how much the unquestioned assumption of an implacable technologically advanced enemy ramified throughout the practices and presuppositions of American science. The immediate fallout was the contraction of direct defense-related R&D after 1988, while nondefense expenditures failed to take up the slack, as illustrated in figure 3.

However, shrinking Federal research budgets were just the tip of the iceberg that had punctured the supposedly unsinkable vessel of Cold War scientific research.

It has become increasingly apparent that the national security imperative had stood as the Maginot line of defense of the universities against many forms of encroachment originating from disgruntled elements of its clientele, since its collapse heralded an array of incursions upon universities. Soon after the fall of the Wall, universities came under attack for their lax accounting standards and supposed abuse of indirect cost provisions that had been forged back during WWII. Various high-profile fraud accusations were lent some credence by hastily constituted government inquisitions; and periodic eruptions of litigation made it seem as though universities were not quite altogether capable of keeping their own houses in order. After years of quiescence, the quality and amount of teaching at major research universities came in for scathing criticism, as did the professorate in general. Admittedly, there was some basis in fact for the upsurge of dissatisfaction, since the universities had reacted to the decline in federal subsidies as a signal that the combined teaching-and-research model was no longer being supported by the government, and responded to their own fiscal crisis by resorting to "gypsies," part-time adjuncts and other improvised non-tenured job categories of "unfaculty" as a means of maintaining their position in the research sweepstakes while simultaneously offering the accustomed broad array of coursework. But the irony of the attack on the professorate was that it came just as the job description for that profession was being re-engineered within the new regime.

Reconfigured career paths of laboratory scientists in particular has induced a more collaborative and hierarchical social structure of research. Growth areas of the university are no longer found in the traditional departmental structures, but rather in interdisciplinary research units, often constituted around a particular applied research topic and outside corporate funding. The phenomenon of "research parks" is merely the geographical manifestation of this disestablishmentarian movement. But as research increasingly assumes a privatized character, so too do the teaching functions of the university. A very important development is the increasing use of computers and the Internet to pursue the thorough-going automation of university teaching, under the rubric of "distance learning" (Schiller, 1999, chap.4). Controversial issues of intellectual property, once only the concern of the research scientist, have thus now invaded the classroom. The commoditization of the act of teaching has proceeded apace since his article, with firms such as Blackboard, IntraLearn and WebCT supplying software to facilitate distribution of the courses outside of their universities of origin, and the Department of Defense chiming in with its own plan for standardization of Internet instruction, the "Advanced Distributed Learning Initiative" (Carr, 2000). The alliance of new technologies and a new entrepreneurial culture have given rise to all sorts of automated teaching curricula subject to the privatization frenzy of IPOs and commercial takeover bids, which themselves serve to further fragment the university into profit centers and loss leaders.

Academic career paths have themselves been downgraded, if not deskilled, with the standard career in the laboratory sciences encompassing two or even three "postdoctoral" positions prior to attaining the assistant professorship that aspirants used to count on right out of graduate school under the previous regime. Since the very construct of "academic freedom" had itself been a key structural aspect of the Cold War regime, it was only a matter of time before the practice of granting academic tenure itself came under attack, and the "academic freedom" defense of self-directed inquiry grew to sound increasingly like an exorbitant luxury, if not a hollow catechism. Some universities sought to close down entire academic departments, and the sciences were not always exempted. Centrifugal forces began to separate out what Federal largesse had held together for four decades.

Moreover, it was not only the universities that were experiencing an unanticipated gale of creative destruction. Many sectors of corporate America also initially felt blindsided by the collapse of the Cold War system. The industries that had most benefited from the stealth industrial policy were the first to feel the chill winds of restructuring. Close on the heels of the fall of the Wall, many of the corporations with the most illustrious in-house R&D units decided upon the withdrawal of military subsidy they could no longer afford such forward-looking subsidiaries, and sought to either downsize the units or spin them off as free-standing firms. The alternative model for R&D has increasingly been to seek collaborations where directed research can be outsourced from the corporation. It has been estimated that in 1997 roughly 10% of all US companies outsource some portion of their R&D, and in research-intensive industries like pharmaceuticals, the proportion is more like 37% (Borchardt, 2000). Some of these "contract research organizations [CROs]" benefiting from the R&D spinoff can be found in those research parks abutting American universities; but increasingly, CROs are to be found in areas like the former Soviet Union or Ireland, where labor costs are significantly lower and the temporary character of employment for research staff is much more accepted.

Not unexpectedly, the other salient aspects of the stealth industrial policy were also summarily reversed (Branscomb et al, 1999). Antitrust policy veered in a much more lenient direction; while legal strictures on intellectual property rights grew much more stringent. Corporate legal staffs thrived while research units were downsized. In the new regime, the watchword was "privatization" of functions and entities that had previously enjoyed governmental subsidy; what this meant for corporate R&D is that it could be separated out as a modular profit center subject to thorough-going restructuring, as it was subordinated to the competitive strategic position of the new multinational firm. The mutual interaction of the induced economic vulnerability of the universities with the novel corporate drive to reinvent contract research, itself the outcome of a change in Federal government policy, has resulted in a new science policy for a global privatized economy of information.

As in all economic formations, there are golden opportunities, and there are raging disasters. One opportunity has been for academic entrepreneurs to pursue these franchised research projects, and to use them as leverage to change the governance structures of the structurally weakened university. Academic disciplinary boundaries, and even the budgeting and planning functions of deans and administrators, are eroded by what are essentially commercial deals being negotiated by laboratory heads and directors of research institutes for long-term funding. University-industry research parks are merely the latest topological manifestation, and part-time faculty CEOs a social innovation, of the structural makeover of the university as a site of privatized research capacity. But these novel sources of support of science are not an unmixed blessing. Not only do tangled commercial/academic contretemps surface with increasing frequency; once the research function has been uniformly privatized throughout academia, it will become increasingly impossible to insist upon any residual academic control over the conduct of the research, as illustrated by the 1998 agreement between the University of California-Berkeley and Novartis, giving it an active veto in the research committee of the Department of Plant and Microbial Biology (Press & Washburn, 2000). This loss of independence is already endemic in the lucrative area of clinical drug trials for pharmaceutical concerns. Drug companies, finding that clinical trials of new drugs both take too long and are too expensive when conducted at university hospitals, have turned to entrepreneurs who recruit general practice physicians to themselves recruit their patients as clinical subjects and to conduct basic experimental protocols (Eichenwald & Kolata, 1999). Beyond the thorny ethical issues raised by the perverted incentive structures in this system, it should stand as testimony to the fact that, if universities privatize research, then there is no guarantee they will long remain the low-cost provider of corporate research services. In an increasingly globalized setting, short-term research contracts can be as volatile as short-term capital flows.

With the rise of the globalized privatization regime, we witness the increased importance of the economic value(s) in and of science and a growing need to a philosophical perspective on this development. There are several philosophical starting points for economics of science. Many authors dealing with an economics of science begin by trying to distinguish the character of "pure" science from other pursuits, such as technological development, humanistic research, or even play. The problem of distinguishing "pure" from "applied" science and separating science from technology has been the bane of science policy since its inception (Forman, 1997; Reingold, 1995; Ziman, 1994). However, the pure/applied distinction is becoming less relevant, for reasons outlined in the previous section, and even the traditional separation between individual sciences is becoming less important. Hence, older attempts to maintain and reinforce these boundaries grow irrelevant. In addition, the stress on "ends" or "intentions" of the participants has not been a very fruitful point of departure, if only because it does such a poor job of describing the personal experiences of scientists and the major themes that divide scholars concerning the legitimate use of economic terminology to describe science.

Instead, we will suggest several alternative ways in which the literature can be divided. For instance, it seems more promising to distinguish the various kinds of social structures that reputedly characterize scientific endeavor versus the kinds of social structures characteristic of a market economy. This mode of approach is one of the hallmarks of the so-called new economics of science: "the new economics of science has the two-fold ambition of (1) exposing the underlying logic of the salient institutions of science and (2) examining implications of those differentiating features for the efficiency of economic resource allocation" (Dasgupta & David, 1994, p. 492). By removing the question of the "ends of science" to a more remote philosophical location, one can more readily see how different economic orientations could shape different modes of evaluation of the efficacy of the social structures of science.

In the history of science studies, one can distinguish five possible positions that could be the starting-point for evaluation of science/economy linkages. First, science is an entirely self-sufficient social formation that operates according to its own autonomous principles or norms antithetical to the market. This was largely the orientation of Mertonian sociology of science (Hagstrom, 1965; Merton, 1973; Storer, 1966). Second, science and markets are two classes of self-organizing social formations that exhibit some vague metaphorical similarities, yet operate largely independently. In contrast to markets, science should not be justified on utilitarian grounds, however. This would be the position of Michael Polanyi (1962; 1969), for instance. Polanyi would argue that we are still in the early stages of scientific enlightenment and that science budgets still need to be protected from public scrutiny. Although often regarded as critical of science, many constructivist sociologists of science fall into this category as well (Knorr-Cetina, 1982). This would also seem to be the position of many scientists who have ventured into larger science policy debates (Gross & Levitt, 1994).

Third, science and markets are two distinct classes of social activities that depend upon one another in a symbiotic fashion. While science is conceived as a set of primarily non-market structures that organize the learning and validation of knowledge of the natural and social worlds through some sort of "reputational credit" or "cultural capital," markets control the dissemination and uses of knowledge. This would be the position of Paul Feyerabend (1978; 1987), for example. Feyerabend would argue that the time is long overdue for science to be made the subject of complete public accountability, outside of internal reward mechanisms. Furthermore, the tradition that makes a sharp distinction between pure science and technology, and then locates the "source" of economic growth in technological change, discussed briefly in the introduction, also falls into this category. This would also seem to be the position of the major protagonists of a "new economics of science," Partha Dasgupta and Paul David (1987; 1994). Fourth, markets are the general paradigm for all modern social organization, and therefore science is then just one special case of this generic social structure. Many cognitive questions can be reduced to problems of individual constrained optimal choice. Philosophical writers who adopt the metaphor of a "marketplace of ideas" (Kitcher, 1993; Rescher, 1989) fit into this category, as does much of what has appeared as the "old economics of science" (Diamond, 1988) or the "economics of information" (Babe, 1994; Boyle, 1996), especially in economics journals. Finally, both science and markets are distinct social activities that display no intrinsic essential attributes, but have evolved in parallel historically from their individualist handicraft origins into large-scale multi-actor hierarchical entities. This would be the position of the "Social Relations of Science" movement of the 1930s (Bernal, 1939) to the "social epistemology" movement of today (Fuller, 1988; 1992).

Obviously, a fuller elaboration of the possible interpretations of markets and science/economy linkages is possible. In addition, once such linkages are established, the question arises as to the characterization of the "goods" that are "produced" and "consumed" in science, to which we turn now.

The more explicit opinions concerning science/economy linkages discussed in the previous subsection are intimately connected with more implicit views about the characterization of knowledge. Discussions of the economic value(s) in and of science inspired by neoclassical economics often slide from science to knowledge to information. In one approach, for instance, self-interested scientists guide the scientific economy to an efficient equilibrium of supply and demand of information, with no need for science funding (Diamond, 1988; 1994; 1996; Stigler, 1961). In another interpretation, however, science needs to be funded because of the underproduction that goes along with its public good character (Arrow, 1962; Nelson, 1959).

Since much of economics is about exchange, the question arises what is exchanged in science. For economics of science, the idea of a marketplace of ideas and the attendant commodification of knowledge raise some thorny issues (Boyle, 1996). On the one hand, efficient market exchange requires free information. On the other hand, optimal individual incentives require the commodification of information. If science is equated with knowledge and if knowledge is equated with information, there may be no trade, or no science, in efficient scientific markets. In addition, the characterization of knowledge underlies one of the most contentious issues in an economics of science: what is the relationship of the science analyst to the working scientist; and (if relevant) what is the relation of the scientist to the actors who are the object of scientific study (Sent, 1998)?

Furthermore, there are at least five philosophical issues that an economics of science confronts, dealing with what knowledge is, how it is produced and how it might be automated. First, there may be a stress on either individual cognition or social structures of cognition and interpretation. Second, knowledge may be treated either as context-free, global, and universal or as context-dependent, local, and situational. Third, knowledge may be considered additively cumulative, or subsequent knowledge may alter and revalue prior knowledge. Fourth, knowledge of the natural world may be of a different character and display a different accessibility than knowledge of the social world. Finally, knowledge may be considered to be either a purely cognitive phenomenon or partly accessible and partly inaccessible to the conscious mind.

During the Cold War, a large amount of science funding revolved around the military, which took the distinction between "pure" and "applied" science for granted and funded "pure" science based on a generally accepted public goods argument. With the end of the Cold War, the decline in military funding, hostility towards government interference, skepticism about the telos of science, questions about the accountability of science, the push to develop connections between business and academia, and the like, it is no surprise to encounter more and more discussions of the economic value(s) of and in science.

This paper has suggested ways for economists, historians, philosophers, sociologists of science and representatives of the sciences to come to some understanding as to what might be useful in notions of the economics of science. The emerging field of economics of science can only be advanced by discussing the various approaches in a systematic, comparative and integrative manner. To this effect, the paper has given a historical and philosophical background and surveyed some theoretical and empirical approaches. The next step is to pursue these in more detail.

We believe the future lies instead in a discussion of the economic value(s) in and of science that takes as its mandate dogged confrontation of the big questions facing science in the new century. These would include: What will happen to the university once research and teaching are spun off as separate privatized self-contained endeavors? When most research is being carried out under the auspices of some corporate agency or funding agreement, how will people come to factor that into their own assessment of the validity or dubiousness of the published account? Once scientific journals all get transformed into electronic archives, what will that do to the structure of the professional organizations whose raisons d’être were the care and maintenance of said erstwhile journals? When certain scientific specialties such as particle physics or algebraic topology (or the history of economic thought!) fail to find their corporate patrons, what will happen to them? Can universities continue to support these sad homeless creatures just off the revenues from past patents, or will the successful sciences demand them as their rightful tribute? Indeed, can universities survive as the repositories for everything which corporations have considered and found useless? What is the role of governmental funding agencies such as the National Science Foundation or the National Institutes of Health in the chastened circumstances new globalized privatized regime? While the history we have proffered does not give grounds for boundless confidence, we do feel justified in closing with the observation that days of science portrayed as a black-boxed "marketplace of ideas" are over.