Questions About Cancer? 1-800-4-CANCER

Carey Lecture - AAAS Policy Meeting

Harold Varmus
President and Chief Executive Officer
Memorial Sloan-Kettering Cancer Center

APRIL 22, 2004

(as delivered)

Science, Government, and the Public Interest

For over six years, while serving as Director of the NIH, I was often asked to speak at S and T policy meetings, not unlike this one. The obligation that I felt, and duly exercised, was to give my news and my views on hard core policy issues---budgets (especially), legislation, scientific priorities, technology transfer, and many other topics.

Today the occasion is very different---an honor for which I am grateful, not a duty to which I must respond. And today I represent only myself, not a large agency, the grantees dependent on it, or the Administration that appointed me. So I thought it would be enjoyable, and perhaps even useful, to address something more fundamental to contemporary life in American science than the "policies de jour." I have chosen to do this by speaking about several aspects of a central relationship in modern American life, the intersection of science and government. I would like to reflect on the importance of that relationship and its limitations, by describing what I view as some current dangers to its continued health and some unrealized opportunities for synergy and progress.

The relationship between science and government has not always been as richly textured as it is now and was in the second half of the twentieth century. Historically, science, like philosophy and the arts, was generally either pursued as an avocation by the wealthy or financed by wealthy patrons, who might or might not also have political power. Sometimes the patron was an institution, the church or even the state, in the form of dukes or royalty. More recently, especially in the first half of the 20th century, it was often one of the new philanthropic foundations. Following the industrial revolution, as commerce learned to use the fruits of science in the 19th and 20th centuries, increasing levels of support also came from the industrial sector, commonly (but not always) with the intention of producing something of value to the investor, not necessarily knowledge for the general public.

The extraordinary characteristic of governmental involvement in science is its most prominent purpose: to create knowledge that advances public welfare---commonly the welfare of the nation but often global welfare as well. To achieve this goal, science and government need to be mutually supportive. In its simplest form, the government pays scientists to make discoveries (also known as "public goods") that lead among other things to practical inventions, economic prosperity, and better health---all of which are in the public interest. The relationship depends on trust between the two parties and allegiance to rules that are often unspoken and not legislated, but worth trying to state here.

To get what it needs and wants from scientists, the government needs to ensure reasonable levels of financial support for scientific work; organize agencies that can deploy the funds in a fair, equitable, and productive fashion; pay attention to the supply of scientific workers; observe sympathetically the physical and functional components of the scientific infrastructure (known as Facilities and Administration or F & A); listen to the advice scientists offer to the government on many issues; and exercise wise oversight---neither careless nor draconian---of the agencies and institutions that spend public money, keeping an eye on priorities, on the distribution and use of scientific findings, and on scientific integrity and fiscal accountability.

In turn, the scientific community needs to honor the government's fiscal commitment. This is achieved by individuals and their institutions through the hard work of doing science; communication of research findings; the education and training of new scientists; the pursuit of useful applications of new knowledge; an enthusiasm for providing non-partisan advice about scientific knowledge that informs policy making; and a willingness to be subjected to competitive review by peers and to administrative and legislative oversight by government.

This relationship clearly covers a lot of territory, and I have chosen six items on the map of that terrain that I want to discuss today.

A Path Not Taken Tonight

But first I'd like to briefly mention and set aside one important item that is not on today's list. No one who follows the news about science in the US can be oblivious to the widespread concerns recently voiced by many senior scientists about the methods being used by the current Administration to obtain, evaluate, and portray the advice it receives from the scientific community. Advice from scientists has played a central part in the development of policy by US government throughout our history, especially since the establishment of the National Academy of Sciences in 1863, and even more since the establishment of the office of the President's science advisor . Current concerns about the selection of some of those who are asked to give advice and about the treatment of advisory reports have been elaborated in a widely disseminated letter, which I and about sixty colleagues have signed. This letter was accompanied by a detailed report by the Union of Concerned Scientists, presenting several specific instances of what we view as inappropriate behavior.

This is not the place to air disputes about current processes. But the matter illustrates the delicate balance that exists in the relationship between science and government. Some have argued that we have taken unnecessary political risks for the scientific community by releasing such a critical report at the start of an election year, since we are dependent on governmental support for so many things. I believe the signers recognize those risks and have assumed them because of the importance of the principles at stake. For regardless of what public policies are ultimately shaped by an administration, it is essential that the best available scientific information be used in policy formation. This is a strategically vital intersection of government and science, and we should not allow what we perceive as widespread distortion of it to go unchallenged, even in an election year.


Instead let's start by talking about money--- the indisputable glue in the relationship between government and science. At this time in our history, the vast majority of scientific work performed in what is essentially the public sector---in academia and government---is absolutely dependent on the availability of government funds. Furthermore, the entry of smart people into scientific careers is dependent on the perception that funds for research will be available in the future. Regardless of how that money is divided among the agencies that support science and then divided again among the activities of each agency, it is essential that the national treasury be equipped to sustain the vitality of American science.

The darkest cloud on the horizon of the marriage of government and science is the rapidly growing budget deficit---a product of expanding costs of mandatory programs, an expensive war in Iraq, and dramatic reductions in revenues by ill-timed tax cuts.

In an amazingly rich country like ours, with an annual gross product of about $10 trillions, it would seem a simple matter to have enough set aside to insure that all the major sciences are not just maintained, but growing at the kinds of rates we consider healthy for other components of our economy. This is especially so when leading economists agree that public investments in science have a rate of return unmatched in any other area. But the Federal deficit that now yawns before us threatens essentially all sciences with not just an absence of growth, but possibly a reduction in support over the next several years.

Consider the consequences and you will recognize the extent to which the government-science partnership pervades our lives and our economy: less knowledge to advance the public's interest and support industrial innovation; the decay of universities; fewer US students entering science or foreign scientists coming here to do it; a loss in international prestige.

It is time to reassert the depth of our relationship and to redefinition its fiscal requirements, continuing to reward accomplishment and preserve competition, while providing stability. It is simply true that the scientific community makes a unique plea when it raises concerns about tax reduction and deficit growth. I recently heard the distinguished economist Joe Stiglitz of Columbia U say that he'd be a lot more comfortable about the deficit if it were a consequence of expanded investments in science. I wish someone other than my colleagues had been listening!


After finances, nothing is more important for the future of science than the quality of the cohort of future scientists. The availability of long term resources for the scientific enterprise is, of course, among the factors required to insure a steady supply of new talent for science in the public domain.

A second factor is the training environment—the schools and universities and lab-based programs that we use to interest students in science and to teach them to become good scientists. We continue to do a good job in many fields, like biology and computer science, with what David Goodstein at CalTech has called mining diamonds---finding gems and polishing them---but we don't do so well in all fields, especially engineering and mathematics. And test scores show that we are mediocre in teaching the scientifically ungifted or the disinclined, those who eventually become the general public. (But this is another story for another time, and for someone more knowledgeable about education than I am.)

The third factor is our capacity to attract talented and ambitious young scientists from abroad. This is true especially in those fields to which few of our own students have been attracted in recent years, notably math and engineering. But even the biological sciences are increasingly dependent on trainees and young faculty who were born elsewhere. In my own history of training over 100 students and post-doctoral fellows, about half originated abroad, from over twenty different countries on almost every continent; about half of these have remained in this country in long-term jobs. My experience is far from unusual, in my field or others.

Enlightened immigration is a long-standing source of both pride and outstanding scientists. (Bruce Alberts reported in his annual Presidential Address to the NAS this week that fully a quarter of current members of the NAS were born abroad.) But we also have a history of ambiguous attitudes towards immigration and towards visitors too, fueled by fears of the financially burdensome, the infected or infirm, the politically dissident, or, most recently, raw terrorists. Such fears need to be balanced by our continued need and desire to be a magnet for talent and ambition, as well as a haven for the oppressed from around the world.

But there are signs that practices affecting visas for students and scientific visitors have recently veered out of balance. The GAO reported last year that the average time for issuing visas is longer than two months; there are many examples of much longer delays, although also the State Department has claimed that visa delays have been reduced recently in response to such complaints. Still, students and fellows report treatment that is annoying, insulting, and humiliating; this sends a message.

One possible consequence is described in a recent report from the Council of Graduate Schools: a decline this year of about one-third in applications for graduate training, especially from China, Korea, India and several other countries. Moreover, there is anecdotal evidence that they may be choosing to go to other English-speaking countries, such as Canada and Australia, or European countries. In his Presidential Address, Bruce Alberts commented that "we will lose something that is irreplaceable if this dreadful situation persists much longer." And in a passionate column on this topic in today's New York Times, Tom Friedman describes the anxieties that recent practices are provoking in technology-based industries that are dependent on foreign visitors and immigrants---on what Friedman calls our "ability to skim the cream off the first-round intellectual draft"--- to maintain a competitive edge against China, Japan, India and other Asian countries, as well as traditional European rivals.

Certainly, the scientific community has a responsibility to recognize legitimate concerns about terrorism; they will and should have an effect on procedures and policies that govern travel and immigration. But the government must formulate a reasonable defense; fears of terrorism must not be allowed to erode our ability to attract talent to our shores. Few things would be as short-sighted for the continued prosperity of science in America.

As Tom Friedman says in today's column: "We have got to get our focus back in balance, not to mention our budget. We can't wage war on income taxes and terrorism and a war for innovation at the same time."


The contract between government and science has worked well in the US in large part because the scientific community has made an enormous commitment to police the quality of grants that are awarded and papers that are published in leading journals through expert peer review. While understanding its imperfections, we in the scientific community also assume that the near-sanctity of peer review is widely accepted. We are prone to worry about fine-tuning the system---optimizing review groups and reevaluating criteria---so it is easy for forget some simple truths: how much time our scientists devote to reviewing the work of their colleagues; how important expert review has been for establishing the credibility of public science with the executive and legislative branches of government; and how different things are in most other countries that spend considerable amounts on science but then entrust the distribution of funds to other mechanisms---bureaucrats who never learned science, superannuated scientists who have forgotten what they once knew, life-long entitlements, and university administrators.

Like anything that requires so much work, peer review is fragile, especially the time-engulfing process of reviewing grant applications. Two things threaten it. One is a decline in funding that can lower success rates to the point at which the review mechanism cannot make credible decisions between what should and should not be supported. In this situation, review criteria are corrupted----innovation gives way to mere feasibility---and older established scientists are given opportunities that belong to younger untested ones. This happened on a broad scale at NIH in the early 1990's and, despite the recent doubling of the NIH budget, it could happen again. Policy analysts have shown that budget deficits over the next few years drive us back to success rates reminiscent of the early 90's.

The other threat is more immediate and more distressing: an effort to undermine peer review through poorly informed political action. Traditionally, the NIH has used its Advisory Councils, with their several public members, as a way to ensure that the public's interests are represented in the awards process and to defend against inappropriate intrusions. On July 11, 2003, Mr. Toomey of Pennsylvania rose on the floor of the House of Representatives to propose an unusual amendment to the NIH spending bill: four NIH grants would be stripped of their funds because he had determined from the abstracts describing the awards that they were inappropriate for funding. Fortunately, the allies of science came to our aid, and the Toomey Amendment was defeated---but by a mere two votes.

This worrisome episode tells us that there is a festering wound in the relationship of science and government. We as scientists and the science agencies need to describe our review processes with greater clarity and in wider venues to insure an appreciation of them.


The quality of science in the US depends substantially on our history as a basically secular country. We were founded, it is true, by people of Christian denominations, often escaping religious persecution. The experience of religious persecution may explain why faith is much less prominent than concepts of liberty, human rights, and the organization of government, one with balanced powers, in our founding documents---the Declaration of Independence, the Constitution, and The Federalist papers. Yet, ironically, as recent immigration trends have made our country much more diverse culturally, ethnically, and spiritually, we have not become more securely secular. Instead, an increasingly dogmatic faith-based element has invaded government and politics, undermining the evidence-based approaches to problems that most scientists would like their governments to use. In crucial situations, this can produce important mistakes with disastrous consequences---even well beyond the usual confines of science, as in trying to find "weapons of mass destruction" that we know must exist in "evil" countries, rather than looking for evidence that they do.

Arguments of the sort used by Mr. Toomey, based on the contention that spending tax dollars to do certain research might offend the sensibilities or beliefs of taxpayers, threaten to replace another kind of moral argument based on the idea that spending tax dollars to do certain research might produce benefits, especially health benefits, for many diverse people in our complex society. There are now many examples of this kind of reasoning; I will mention just two more. First, the strong tilt in this Administration towards abstinence programs and away from more realistic programs that use contraceptive devises, like condoms, to prevent the spread of HIV and other sexually transmitted agents, here and even abroad. And, second, the policies that now govern stem cell and human embryo research, that give heavy weight to the moral rights of a tiny cluster of undifferentiated cells at the expense of full fledged adult citizens who could benefit from such research. These policies are now driving most of this work to the private sector, to a few wealthy non-profit institutions, and, most troubling for our nation's future in science, to new players like South Korea or to our traditional competitors like Great Britain.

This is not an argument to deny ethical considerations their role in deciding how to conduct government-supported science. But this role needs to be exercised in the context of a balanced, non-doctrinaire approach in which the ethical consequences of research are also fully considered for application in an increasingly pluralistic society.


The US government generally understands its role as the major supporter of the scientific work that benefits our citizens and our economy. But it has been slower to understand the global impact of the public goods that science generates and the beneficial effects---many in our self-interest---of US support for science in other countries, including the poorest countries.

We now achieve our best effects internationally by training foreigners who return home and also by permitting widespread use of knowledge that we produce here and place in the public domain. But I envision a much more extensive program with more expansive goals. I believe that we can "globalize science" in a way that builds sounder societies, links scientific communities (and, secondarily, their nations), and produces knowledge with regional or national, as well as world-wide, importance.

At the moment, these ambitions are pursued largely by NGO's, foundations, and a few northern European countries more intent than we have been on fulfilling development goals defined by the United Nations. For relatively small amounts of money, our government can build upon the work that has been done by others to develop much more active scientific communities in many other countries. For instance, the expanded interests in global health research that have been promoted by initiatives at the Gates Foundation, especially the recent $200 million Grand Challenges in Global Health project, should provide incentives for the government to expand its own efforts in this important research arena. Perhaps it is time to convert the small backwater of international activity at the NIH, the FIC, into a legitimate, well-funded, full-fledged National Institute for Global Health. (Note to Varmus watchers: this is not a renunciation of my cry for no more components at the NIH---it would keep the number the same---but a reflection of a serious need for more science to confront problems in the developing world.)

My enthusiasm for building science in the developing world is based on several things---the belief that science can improve lives in those countries; the fact that the opportunities to become a scientist are severely limited there (Ismael Serageldin has noted that the chances of becoming a scientist in a poor country are only one percent as great as the chances of becoming one in a rich country); the desire to counter the damage we are doing to our international reputation for beneficent leadership by our actions in Iraq and elsewhere; and the experience of seeing American dollars for science at work in one of the poorest places in the world, Bamako, Mali. There, efforts by USAID and NIAID have created a strong Center for Malaria Research and Training that is a source of local health improvements, recruitments of Malians to medical science, national pride, and internationally respected science. We need more Bamako's.

Over the past couple of years, I have been urging various funding agencies, within and outside governments, to support a Global Science Corps that would place trained scientists from the developed world---especially people at early or late career transitions or during sabbaticals---in high quality laboratories in developing countries to provide technical expertise and teach, while learning themselves about new scientific problems. This effort is now being run by the Science Initiatives Group and gaining some financial and governmental encouragement. This is a remarkable opportunity for governments and scientists to advance all dimensions of the public interest---easing international tensions while advancing discovery of useful knowledge---for relatively modest sums.


Our government is spending billions of the public's tax dollars to generate knowledge meant to be public goods. Yet the scientific community has not done as much as is now possible to optimize the dissemination, storage, retrieval, and use of that knowledge. The government has an obvious vested interest, on behalf of its citizens, to see this happen, But too it has not yet moved vigorously to encourage the use of its own technology---the Internet---to make the work it pays for immediately and freely available to everyone, everywhere, stored in and retrieved from digital libraries it has already begun, such as PubMedCentral at the NIH.

This dream of freely accessible public knowledge has been around for a long time, long before the digital age. In 1836, the head of the British Library said: "I want a poor student to have the same means of indulging his learned curiosity, of following his rational pursuits, of consulting the same authorities, of fathoming the most intricate inquiry as the richest man in the kingdom…."

We now have the technical tools to make this vision a reality. The advantages to scientists, to students, and to the growing number of interested citizens everywhere in the world are obvious. What stands in the way?

First, concerns about how we will pay the costs of open access publishing. But these costs will inevitably be less overall than the escalating and increasingly unaffordable costs of the traditional, subscription-based, restrictive model of publishing that we currently use. And government is already covering most of those costs in the US, through grants that pay for laboratory subscriptions, for page and color charges, and, through indirect payments, for libraries.

Second, concerns about the fate of scientific societies that support other worthwhile activities with revenues from their journals. There is no doubt that some societies, those dependent on such revenues, will have to adjust their business plans and obtain more revenue from membership fees, meetings, and other services. The first objective of any scientific society, just like any union or guild, should be to optimize the working conditions for its scientists; surely making the scientific record freely accessible and more usable should be the paramount consideration.

Third, concerns about the survival of revered and expensive-to-produce journals like Science and Nature. But the open access movement is addressed to primary research reports, not to the costly, entertaining, and important "front matter" of these journals---the news, editorials, obituaries, gossip, book reviews, and mini-reviews. The front of the book will be sold for the foreseeable future. And the scientific reports can be provided for free, in digital form only, entirely separately, without significant loss of subscribers. For an organization like the AAAS, dedicated to the public's welfare, this would be doing the right thing.

The government also has a role to play here. Publication should be viewed as part of the cost of doing research (in most fields it is also a very small part of the cost, less than 1%); the government should expect to pay publication costs, even when they are shifted from reader to author, just as they do now; and rapid open dissemination will augment the value of the research by promoting its use as a public good. A firm statement of those principles, including an expectation that science supported by public money will be publicly available, will help cement the resolve of scientists, who know that open access publication should and will happen, to make it happen soon.


Beginning with World War II and the era of FDR, the US government has forged a remarkable partnership with the US scientific community. The generous Federal support of many disciplines in enlightened ways is largely responsible for our advances in health and medicine, our military superiority, our agricultural and industrial productivity, our university system, our leadership in technology, our very prosperity.

But this is both the best and worst of times for this partnership. We continue to lead the world in scientific discovery, our research universities are still unmatched in other countries, new findings and new technologies are bringing different disciplines to intersect in unexpected ways, and public interest in science continues to grow.

Still, as I have described here today, we have a great deal to worry about. Enlarging budget deficits threaten to constrict future budgets for science, immigration policies and practices may deflect new talent to other countries, and inappropriate attacks on peer review and a failure to separate religion from science can undermine the delicate balance between scientific independence and governmental oversight. Furthermore, we may not be moving fast enough to promote science and its uses in the developing world and to provide open access to scientific findings through the Internet.

Effective collaboration between science and government still seems to me one of the few rational ways to find our path in a largely irrational and increasingly dangerous world. I hope that an accounting of these concerns can ultimately help to strengthen our traditional relationship.