The big idea: Proactive, ethical, multi-attribute risk analysis can avoid stalemates between entrenched, extreme views.

The scenario: An energy company chief executive must choose whether to drill for unconventional natural gas at sites in shale, tight sands and coal-bed methane. These three opportunities require hydraulic fracturing or “fracking.” While this technology was first used in 1947, current conditions make fracking increasingly attractive, as it can release gas unobtainable by conventional means. Fracking forces large amounts of chemical-laced water at high pressure through otherwise impermeable geological structures thousands of feet below the surface, forming fractures through which gas molecules can flow. Potentially as much U.S. energy can be released this way as is under Saudi Arabia, increasing domestic reserves dramatically, reducing the price of energy, and adding tremendous value for shareholders.

However, drillers use unknown chemicals with unknown effects on humans and uncertain occurrence of contamination. This has resulted in documentaries and public sentiment that oppose all fracking. Drillers say the practice is safe, suggesting that pipe casings and layers of impervious rock keep fracking water from aquifers. The science is incomplete regarding how likely it is for fracking chemicals to contaminate aquifers and what the morbidity effects on humans drinking these chemicals would be. Huge amounts of water are needed in fracking, sometimes competing with agricultural or commercial uses in dry areas.

Stakeholders have differing objectives: People living in the vicinity of drilling care about reducing contaminants in drinking water. Energy users, citizens and government leaders want less-expensive energy and less dependence on foreign energy supplies. Shareholders of drilling companies are interested in efficient and abundant production of gas. Competing users want less water used in fracking. Stakeholders can become entrenched in extreme positions, ranging from advocating no fracking whatsoever, to the opposite position of unrestricted drilling. Given such opposing positions, a chief executive can be paralyzed.

A decision maker should first rule out alternatives that are unethical, based on principles, rights and moral considerations. Then consider the combined objectives of all stakeholders, specifying attributes that quantify achievement of objectives. For example, a set of attributes might include human mortality, gas reserves developed, value created for shareholders and water usage.

Second, consider the risks explicitly. Risks to life and limb might comprise the product of three uncertainties: probability that an aquifer becomes contaminated, probability distribution for death and morbidity risks of someone drinking that water, and uncertainty regarding the size of the exposed population. A unit of risk might be the micromort, a one in one million chance of one person dying. Realizing no activity is without risk, compare fracking risk with that of other acceptable activities. One micromort is incurred by flying 1,000 miles on a jet or receiving one chest X-ray. Or compare the risk with that incurred in not drilling.

The resolution: Take account of relevant stakeholders. Specify attributes that measure the extent to which their objectives are met. Identify areas of agreement and conflict. Be explicit about risk; estimate probability distributions for uncertainties. If there is greater uncertainty, make wider distributions. Put this in context of other risks, looking for acceptable levels of risk. Explicitly consider rights and principles.

The lesson: Proactively accounting for objectives of all parties, considering the range of potential outcomes for each and comparing them with acceptable risks in other contexts can help to choose sites where drilling can take place, sites where it should not and middle opportunities that will await better information.

Samuel E. Bodily

Bodily is the John Tyler Professor of Business Administration at the University of Virginia’s Darden School of Business.