Doctors, engineers, lawyers, and many other occupations have formal ethics codes that help members of the professions make difficult choices. What about the U.S. intelligence community? There is a formal statement: "Principles of Ethical Conduct," found in Part 1 of Executive Order 12731 of October 17, 1990.a These 14 sentences bind all U.S. federal employees, but these sentences deal primarily with what is usually defined as fiscal mattersproper handling of government funds, nepotism, accepting lavish gifts, and other such ill-advised activities. These sentences can be found posted in the hallways of almost all U.S. federal buildings, including those of the U.S. intelligence community (IC).
As important as these sentences are, they fail to address what we are calling "mission ethics" for intelligence officerswhat you can or cannot, should or should not, do in executing your IC agency's mission. Why should this be of concern?
The IC provides intelligence to U.S. policymakers and military commanders. The decisions they make based on such intelligence can have profound impacts on the U.S., the world, and individual citizens. Intelligence officers are keenly aware of this and strive to do their best, often operating in demanding, complex, and perhaps hazardous situations.
Many intelligence officers have had personal experiences dealing with exacting situations that had difficult ethical ramifications during their careers. Existing fiscal ethics were not germane, and any mission ethics that might have existed may differ by organization (even within an agency) and also may differ as a function of which particular group convened to make a decision. There can be significantly different ethical considerations between tactical and strategic missions.
Most IC employees are dedicated, honorable, and ethical public servants. But they function in secrecy and have powera combination that is fraught with temptation. Some may have to suborn citizens of other nations or the agents of foreign governments. Some may perform operations that would be illegal if done against their citizens. Others may make decisions that can have significant unforeseen unintended consequences.
Several of us who have been career intelligence officers contend that we could have benefited greatly from a set of mission ethics studied and deliberated in training classes using case studies, away from our demanding work environment. Such study and discussion would enable contemplation, reflection, and internalization that could help an officer in the heat of action when others might be overwhelmed with expediency or emotion, or in other situations when perhaps faced with unethical opportunists (there are always a few such in any organization).
Ethics will not solve all problems or make decisions easy, but should help reduce the number of bad decisions that might be regretted later.
Could such a code be devised that would truly be useful? In October 2005, six IC employees and two corporate employees met to discuss this issue. By the end of the day we had decided devising a code was possible and that it would be well worth the effort.
The intelligence organizations of concern to us typically report to nation-states that have differing legal structures, cultures, and value systems. We restricted our attention to the U.S. intelligence community, which functions as an agent of the U.S government, a government designed to be responsive to its citizens and their values. Differing contexts would lead to differing ethics codes.
We agreed on several points:
We explored other ethics codes (including medical, legal, security, and military) as models, which helped us structure our code.
An unclassified code has many benefits. It keeps the code from being too detailed or focused on arcane intelligence matters such as sources and methods. It might also offer some protection against public outrage when classified actions become publicly known (and at times they will). If the citizenry accepted the published code language and the exposed action is in line with the public language, it could help citizens understand the rationale for the action and lessen adverse reactions or possibly offer an opportunity to further refine the language (and constraints on future actions) to be more in line with national values that may change over time.
Also, being public may actually give the ethics code more "teeth" because employees will know that not just their bosses could be sitting in judgment on their actions. This fact might encourage junior employees to resist (however gently) bosses who might be ordering questionable actions, as well as reducing the number of times such bosses might be tempted to order such actions.
We knew the task of crafting actionable language of value would be difficult; we attempted a draft as an exercise to judge the difficulty. We explored other ethics codes (including medical, legal, security, and military) as models, which helped us structure our code. We started from fundamental principles. Our oath as federal employees to support the U.S. Constitution was a key factor and helped prioritize other allegiances we had, such as to bosses, organizations, and fellow citizens. For example, we decided the required oath-of-office we all take to support the U.S. Constitution also requires us to first serve our fellow citizenssince the Constitution's first words are: "We the People of the United States, in order to...".
We concluded that writing a code was possible, but not necessarily easy. We expanded the effort in December 2006 to a closed, invitation-only Internet discussion group of approximately 50 people (approximately one-third with substantial IC experience, one-third professional ethicists, and one-third other). As expected, the ensuing discussion was difficult workinvolving lengthy sessions on phrasing to clearly and properly capture the intended meaning succinctly, even for nuanced concepts.
By late 2008 we felt we had a reasonable draft (see the accompanying sidebar; note that in the draft sidebar content we have intentionally retained some minor structural flaws so that anyone actually considering adopting it would face at least minor edits, which could help trigger the deeper reflection we believe is needed by any organization considering adoption). Producing the draft drove home the point that the real value to all of us was the work in creating it, the forced reflection and reconsideration of beliefs, not the final text. We submit that the most effective ethics training will be achieved if officers engage in debating and arguing the points of the proposed code. Nevertheless, we think the text can be of great help to an employee trying to do the right thing in a specific intelligence circumstance; what would it demand of the employee? We did not try for precise wording that would cover all circumstances; instead, we strove to capture the intent and let it foster discussion, deliberation, and debate that would help people internalize the code.
The current draft is not perfect and certainly can be improved, but those of us who have wrestled with this think it is a good enough draft for the U.S. IC to seriously consider as a basis for further work that would lead to the adoption of such a code of mission ethics.
We invite the U.S. IC to undertake such an effort and hope that intelligence communities of other nations might consider it. Do not simply copy our words, but craft your own, suitable to you and created through concerted effort and discussion. Then develop training that will help you internalize them as a basis for action.
Share your code with the public: we think it will benefit you and gain support for, and acceptance of, your activities on behalf of your nation.
a. See http://www.usoge.gov/laws_regs/exec_orders/eo12731.pdf
Please send comments to: [email protected].
DOI: http://doi.acm.org/10.1145/1536616.1536630
The Digital Library is published by the Association for Computing Machinery. Copyright © 2009 ACM, Inc.
Nice idea, but totally unrealistic.
E.g. the CIA would have to fire at least 50% of it's employees because of misconduct. And the NSA would have to be closed.
Displaying 1 comment