8

Science, Heal Thyself

NEW YORK – Science may be humankind’s greatest success as a species. Thanks to the scientific revolution that began in the seventeenth century, humans today enjoy instant communication, rapid transportation, a rich and diverse diet, and effective prevention and treatment for once-fatal illnesses. Moreover, science is humanity’s best hope for addressing such existential threats as climate change, emerging pathogens, extra-terrestrial bolides, and a burgeoning population.

But the scientific enterprise is under threat from both external and internal forces. Now the scientific community must use its capacity for self-correction – based on new information, discoveries, experiences, and ideas (the stuff of scientific progress for centuries) – to address these threats.

A major hindrance to scientific progress is the increasing scarcity of research funding – a trend that has been exacerbated by the global economic crisis. Uncertain funding prospects not only discourage scientists from pursuing risky or undirected lines of research that could lead to crucial discoveries; they also make it more difficult to recruit the best and brightest for scientific careers, especially given the extensive training and specialization that such careers require.

Furthermore, leaders from across the political spectrum are questioning scientifically-established principles – such as anthropogenic climate change, evolution, and the benefits of vaccination – with no scientific basis. At best, such statements serve as a distraction from important issues; at worst, they distort public policy. Although such threats are outside of scientists’ direct control, improved communication with political leaders and the public could help to reduce misinformation and bolster confidence in science.

But the field’s credibility is also being undermined from within, by the growing prevalence of scientific misconduct – reflected in a recent spate of retracted scientific publications – and an increasingly unbalanced scientific workforce that faces perverse incentives. Although the vast majority of scientists adhere to the highest standards of integrity, the corrosive effects of dishonest or irreproducible research on science’s credibility cannot be ignored.

The problems are rooted in the field’s incentive structure – a winner-take-all system in which grants, prizes, and other rewards go to those who publish first. While this competitive mentality is not new in science – the seventeenth-century mathematicians Isaac Newton and Gottfried Leibniz spent more than a decade fighting bitterly for credit for the discovery of calculus – it has intensified to the point that it is impeding progress.

Indeed, scientists today are engaged in a hypercompetitive race for funding and prestigious publications that has disconnected their goals from those of the public that they serve. Last year, for example, when C. Glenn Begley and Lee Ellis sought to reproduce 53 “landmark” preclinical cancer studies, they discovered that nearly 90% of the findings could not be reproduced. While the researchers who originally published those studies may have profited from increased funding and recognition, the patients who need new cancer treatments gained nothing.

Moreover, this winner-take-all system fails to account for the fact that scientific work is largely carried out by research teams rather than individuals. As a result, the scientific workforce is beginning to resemble a pyramid scheme: unfair, inefficient, and unsustainable.

The incentives associated with the winner-take-all system encourage cheating – ranging from questionable practices and ethical lapses to outright misconduct. This threatens to create a vicious cycle in which misconduct and sloppy research are rewarded, undermining both the scientific process and its credibility.

The problems are clear. But addressing them requires a prudent strategy that accounts for the structural fragility of the scientific enterprise, in which scientists must complete extensive training, regulation can easily stifle creativity, and funding limitations can substantially delay progress.

Because of this fragility, few countries have been able to establish highly productive scientific enterprises, even though scientific innovation and technological breakthroughs are crucial to a country’s productivity, economic growth, and influence. Given the challenges implicit in establishing and maintaining a robust scientific sector, reform efforts must be undertaken carefully.

At the same time, the reforms must be comprehensive, addressing methodological, cultural, and structural issues. Methodological reforms should include revised training requirements that allow for less specialization, together with improved training in probability and statistics. Scientific culture must be reformed to abandon longstanding practices, such as those that determine how credit is assigned. And structural reforms aimed at balancing the scientific workforce and stabilizing funding are crucial.

Some reforms should be fairly easy to implement. For example, it would not be difficult to win support for improving education in the ethical aspects of scientific research. But other important reforms, such as creating alternatives to the winner-take-all incentive system, will present enormous challenges.

An effective reform strategy should employ the tools of science – specifically, data collection and analysis. More data are needed to understand workforce imbalances, the peer review system, and how the economics of the scientific enterprise influence scientists’ behavior.

Science has been studied by sociologists, historians, and philosophers, but rarely by scientists themselves. Now, with perverse incentives undermining their credibility and hampering research, scientists must take matters into their own hands. Applying the scientific method to the problems of science could be scientists’ best hope for regaining public confidence and reinvigorating the quest for transformative discoveries.