Disclosing Terrorism Threats to Improve Public Safety
Six months after the attacks of September 11, 2001, the Bush administration created a color-coded ranking system to inform the public about terrorist threats. The system’s stated purpose was “to provide a comprehensive and effective means of disseminating information regarding the risk of terrorist acts to Federal, State, and local authorities and to the American people” in order “to inform and facilitate decisions appropriate to different levels of government and to private citizens at home and at work.” The aim was to minimize attacks and their consequences. The system was designed to be flexible and information-based. It provided a framework for communicating the severity of national, local, or sector-specific threats as well as their likely character and timing. The alert system established five color-coded levels of terrorist threat: green = low; blue = guarded; yellow = elevated; orange = high; red = severe. The presidential directive clearly contemplated that alerts would be accompanied by factual information. The directive also made it clear that information was intended to create incentives for action. Each level of alert was meant to trigger threat-specific protective measures by government agencies, private organizations, and individuals. The directive provided that threat levels would reflect both the probability and the gravity of attack and would be reviewed at regular intervals to see if they should be adjusted. The level set was to be based on the degree to which a threat was credible, corroborated, imminent, and grave.
The system provided flexibility. Threat levels could be set for specific geographical areas or for specific industries or facilities. The system provided for case-by-case judgments about whether threat levels would be announced publicly or communicated in a more limited way to emergency officials and other selected audiences. The stated intent was to “share as much information regarding the threat as possible, consistent with the safety of the Nation.”
Once the Department of Homeland Security (DHS) was created in March 2003, the secretary of homeland security was charged with responsibility for setting threat levels, with the advice of the Homeland Security Council. Within the department, the warning system was administered by an undersecretary for information analysis and infrastructure protection.
As of early 2006, the terrorist threat warning level had been raised and lowered seven times, each time from yellow (elevated) to orange (high) and back again. The system generally produced warnings that proved too vague to provide government officials, business managers, or ordinary citizens with incentives to take appropriate protective actions. However, alerts were increasingly specific. On August 2, 2004, the Department of Homeland Security issued a warning concerning three particular facilities: the Prudential building in Newark, New Jersey, and the headquarters of the World Bank and International Monetary Fund in Washington, D.C. On July 7, 2005, when several bombs were detonated in the London subway system, DHS raised the threat level from yellow to orange for mass transit only, though noting that the government had “no specific, credible information” to suggest that an attack in the United States was imminent.
The system worked differently for different audiences. When a decision was made to change the threat level, department officials notified federal, state, and local agencies electronically or by phone and also called chief executives of major corporations, using a secure connection maintained by the Business Roundtable.
DHS also developed channels for communicating threat information without raising the overall threat level. The department issued threat advisories or less urgent information bulletins for specific locales or sectors. Access to these communications was often restricted, however, leaving the public uninformed. Officials explained that such information was shared on a need-to-know basis, since it was often derived from classified sources. A GAO review of a sample of secret threat advisories in 2004 concluded that they contained “actionable information about threats targeting critical national networks, infrastructures, or key assets such as transit systems.”
In practice, however, the terrorist threat warning system remained problematic. Several in-depth evaluations and surveys found that rankings were little used by its intended audiences. The Gilmore Commission, a broad-based congressional commission charged with continuing oversight of domestic responses to terrorism, concluded in 2003 that “[t]he Homeland Security Warning System has become largely marginalized.” On occasion, governors and mayors declined to elevate threat levels or take other federally recommended actions. Public and private groups expressed frustration at the lack of information about the character and location of threats. The commission recommended the creation of a regional alert system featuring specific guidance, as well as training local officials for responses to each threat level.
A report by the nonpartisan Congressional Research Service in 2003 concluded that threat alerts were so vague that the public “might begin to question the authenticity” of threats and therefore ignore them. The report noted that the government “has never explained the sources and quality of the intelligence upon which the threat levels were based.”
Government officials have rarely received information specific enough to act upon. A survey by the General Accounting Office in 2004 found that sixteen of twenty-four federal agencies had received information about elevated threat levels from the media before they received it from homeland security officials. One of the potential strengths of the alert system was that it was constructed to work synergistically with regulatory requirements. Each federal department was required to come up with its own protective measures appropriate to each threat level and to take those actions each time the threat level was raised. However, federal agencies surveyed by the GAO reported that changes from yellow to orange had minimal impact on their practices, since they maintained high levels of security at all times.
State officials, too, reported that they received much of their information about changed threat levels through the media and got little specific information from the government. The GAO survey found that fifteen of forty states learned about threat level changes from the media before they heard from federal officials in at least one instance. State and local officials reported that learning about threats at the same time as the public could carry heavy political costs. State officials also noted that they received conflicting advice from different federal authorities about what actions to take.
The most serious failing of the transparency system has been its lack of meaningful information and guidance. Local officials, always on the front lines in preparing for and responding to disasters, need accurate, specific, and timely information. A report by the minority staff of the Senate Governmental Affairs Committee concluded in 2003 that two years after the attacks on the World Trade Center and the Pentagon, state and local officials had too little information to respond to terrorist attacks. The report noted that effective communication channels still had not been established with state and local officials, so states and localities had no effective way of communicating with one another or of learning fromthe successes ormistakes of others. A June 2004 report by the nonpartisan GAO echoed these themes. It suggested that warnings would be more effective if they were more specific and action-oriented; communicated through multiple methods; included timely notification; and featured specific information on the nature, location, and timing of threats as well as guidance on actions to take in response to threats.
The public remained confused. Information accompanying increases in the threat level often has been vague or irrelevant to the daily activities of most Americans. Most state governments and many local governments have developed their own alert systems which are not necessarily consistent with the federal system. The administration has also sent mixed messages to the public concerning what actions to take. In raising the threat level to orange on September 10, 2002, for example, Secretary Ridge told people to “continue with your plans” but “be wary and be mindful.” In June 2003, Ridge acknowledged that the system needed improvement. “We worry about the credibility of the system...we want to continue to refine it, because we understand it has caused a kind of anxiety.”
Members of Congress from both parties expressed growing impatience with vague and conflicting messages. After the government raised the threat level to orange over the 2003 Christmas holidays and told citizens to be vigilant but continue their daily routines, Christopher Shays (R-Conn.) asked: “Why would the department tell people to do everything they would normally do?...We’re at high risk.” Christopher Cox (R-Calif.), chairman of the House Select Committee on Homeland Security, noted that vague warnings could also cause too much action, citing evidence that groups had canceled field trips and other activities. Senator Frank R. Lautenberg (D-N.J.) noted that “the system may be doing more harm than good.”
Public confusion was reflected in polls. A Hart-Teeter poll sponsored by the Council for Excellence in Government in March 2004 found that 73 percent of those polled were anxious or concerned about terrorism and 34 percent had looked for information about what to do in the event of an attack, but only one person in five was aware of state or local preparedness plans. Earlier Fox News polls found that 78 percent of those responding did not know or said they were not sure what the current threat level was and that 90 percent responded to recent elevation of the threat level by going about their lives as usual. A New York Times poll in October 2004 found that nearly two-thirds of those responding did not have emergency kits prepared and more than two-thirds did not have communication plans.
Philip Zimbardo, the president of the American Psychological Association, suggested that the terrorist threat system had turned the United States into a nation of “worriers, not warriors,” by “forcing citizens to ride an emotional roller coaster without providing any clear instructions on how to soothe their jitters.” He noted that a large body of research suggested that effective safety measures required a credible source, information about the particular event that created a threat, and information about specific actions citizens could take to reduce risks.
In 2011, the National Terrorism Advisory System replaced the color-coded alerts on the grounds that they inspired fear among the public and were too ambiguous to provide useful information. The new system eliminated the color-coded alerts and introduced detailed alerts containing information on whether a threat is "imminent" or "elevated" as well as details on the threat, duration of the alert, and information on the affected areas and actions being taken to protect the public. See a sample alert here. Unlike the old system, which communicated the threat level at airports and on government Web sites, the National Terrorism Advisory System shares alerts via the Department of Homeland security, the media, Twitter and Facebook only if there is a specific threat.
In 2015, the Department of Homeland Security updated the reporting system by adding "bulletins" to the existing alerts. Bulletins were introduced to discuss general trends and current developments regarding terrorism threats that might require additional precautions but do not warrant the issuance of an alert for a specific threat. The first bulletin was issued in December 2015, in the wake of terrorist attacks in Paris and San Bernardino, to discuss the threat posed by self-radicalized groups. As of January 2016, no alerts were ever issued under the reformed system.
Updated February 2016.
This case study is drawn from Full Disclosure, Fung, Graham and Weil, 2007.