< Return To Hearing
Mr. Paul C. Light
April 17, 2002
I am pleased to appear before the Subcommittee today to consider options for strengthening the Office of Homeland Security. I believe this Subcommittee is right on target in asking whether and how the Office might improve the flow of information to and from key decision makers. Although there are many causes of the September 11 tragedy, none was so easily addressed as the failure to collect, interpret, and share information.
I believe the question of information sharing involves two discrete parts. First, is the federal government currently producing the right information? Second, is that information available to the right people at the right time?
The Immigration and Naturalization Service is a perfect case in point. It suffers from a dearth of high-quality information on potential threats to national security, and has serious vulnerabilities in making sure that high-quality information reaches the right people at the right time. Bluntly put, even if the agency were to discover a potential threat, it is not clear that the information would make it to key law enforcement and security officials in time to prevent a tragedy.
We know, for example, that the INS does not have the information technology to track visitors to the United States. The agency simply does not have the right information about who is entering the country under what conditions and for what purpose, nor does it know where current visitors might be located, or whether they have over-stayed their welcome. Although there is nothing the Office of Homeland Security can do at this point to better coordinate, share, or monitor information that does not exist, I believe there are ways that the Office can better control the production of information through expanded authority.
We also know that the INS does not have the technology to share the information it does have with the right people at the right time. We know, for example, that information about the individuals involved in the September 11 attacks on New York City and Washington, D.C., did not make it to the right people at the right time in the Immigration and Naturalization Service and its contractors. Otherwise, one would be hard pressed to explain last month's extraordinary news that ACS, Inc., had mailed visa notices on behalf of Mohamed Atta and Marwan Al-Shehhi. Although one can argue that there was no harm, and, therefore, no foul, the lack of communication up and down the INS hierarchy speaks to the extraordinary problems in sharing information in a timely fashion.
The INS is hardly the only federal agency with problems collecting and sharing the right information. Indeed, recent federal history is replete with examples of breakdowns caused by either bad information or good information ignored, including the security problems at the nation's nuclear weapons laboratories and the taxpayer abuse at the Internal Revenue Service. Looking back with perfect hindsight, we can now see the outlines of a remarkable information breakdown at the core of the September 11 attacks, not the least of which were bits and pieces of evidence from flight schools and simulation centers. As in so many similar cases, from Pearl Harbor to Vietnam, we are likely to discover that the federal government had significant information about the potential threat, but could not put the pieces together in time.
The fact is that the federal government, indeed most organizations, almost always have the information to prevent failure, but not the analysis or communication chains. That was the case in the 1980s with the HUD Scandal and the Savings & Loan debacle, in the 1990s with espionage at the nuclear weapons labs and taxpayer abuse at the Internal Revenue Service, and the early 2000s with the INS. Sadly, I have made a living out of showing the problems associated with communication failures in the federal government. We see the same problems over and over and over again as information gets lost, distorted, mishandled, improperly classified, or mis-interpreted up and down the ponderous federal hierarchy. See the appendix of this testimony for a side-by-side confirmation of my conclusion.
Assessing the Case at Hand
Our task today is not to look back for culprits, but look forward to solutions. Simply asked by the Subcommittee, should the Office of Homeland Security have more power? As I and my colleagues at the Brookings Institution have argued, the answer is "yes." Much as one can credit Gov. Ridge with substantial success in shaping the budgets of the agencies involved in homeland security, his office does not provide the levers that are essential for coordinating, let alone assuring the flow of high-quality information in real time. He should be congratulated for having made the best of a very difficult situation, but should be fully empowered to make sure that the federal government has the information it needs.
Unfortunately, I do not believe Gov. Ridge has the authority to prevent communication failures in the future. Even a cursory review of the Office of Homeland Security organization chart confirms the problem. According to the latest available Yellow Book listing, the Office is structured around an assortment of titles whose primary tasks focus more on outreach than information collection or analysis: (I have numbered the layers within the Office using conventional nomenclature.)
1. Assistant to the President for Homeland Security (Ridge)
2. Deputy Assistant to the President and Deputy Director
3. Deputy Assistant to the President (Communications & Legislative Affairs)
4. Deputy Assistant to the President for Legislative Affairs
5. Special Assistant and Sr Director for Information Integration and Chief Information Officer
6. Special Assistant and Executive Secretary
7. Protection & Prevention Senior Director
8. Special Assistant and Director of Communications
10. Assistant Director for Intergovernmental Affairs
12. Staff Assistant
Much as one my question the problems associated with twelve discrete layers in the Office, and much as one might ask whether the special assistant for information integration and chief information officer is (1) placed high enough in the bureaucratic pecking order, and (b) has too many competitors for access, the more important question at hand is whether Gov. Ridge and his team have the authority needed to assemble the information they need. The answer appears to be "no." At best, information collection and analysis is but one of many competing priorities in the Office of Homeland Security. At worst, it appears to reside in a single unit that has multiple responsibilities, not the least of which is ensuring that the Office itself has adequate computer technology, which is the traditional concern of chief information officer posts across government. At a minimum, I would recommend splitting the information integration and chief information officer posts, while raising the former to the level of deputy assistant to the president.
An Inventory of Authority
More importantly, however, I worry that the Office as currently constructed does not have the authority to implement an aggressive information collection, analysis, and dissemination strategy. Let me suggest the following problems facing the Office today:
1. Production. The Office of Homeland Security does not have the authority to produce information that it deems essential to its planning process. Although it does have a talented, albeit small staff, the Office does not have the internal capacity to investigate or study problems that Gov. Ridge deems essential. To the extent words have meaning, "coordinate" appears 33 times in the president's executive order establishing the Office of Homeland Security, while the word "investigate" appears only once. "Study" and "analyze" do not appear at all.
2. Procurement. The Office of Homeland Security does not have the authority to order or purchase information, whether through private contractors such as the RAND Corporation, which has extraordinary capacity for conducting the kind of exploratory analysis that Gov. Ridge might find particularly useful in anticipating potential threats, or through government intelligence agencies. Nor does the Office of Homeland Security have the dollars to make such purchases. If Gov. Ridge wants a special analysis of trends across or within agency databases, one presumes he must ask.
3. Collection. The Office of Homeland Security appears to have adequate mechanisms for tapping into the federal government's vast inventory of information. The question is whether an office composed of a scant 100 or so individuals, many of whom are dedicated to prevention, outreach, legislative affairs, communications, etc., can hope to swallow from the fire-hydrant of information that current courses through government. I do not see, for example, the capacity to evaluate the quality of information flowing into the Office, nor do I see the ability to reach down into agencies to demand the release of information that already exists. Even more importantly, I do not see the capacity needed to compel the release of information that agencies might or might not know they have.
4. Quality. I do not see firm evidence that the Office of Homeland Security has the internal capacity to assay the quality of information flowing from inside or outside government. At least for the time-being, the Office must rely on its sources to validate information.
5. Analysis and Interpretation. The Office of Homeland Security does not appear to have adequate mechanisms for analyzing information. That might mean, for example, resolving disputes between different portraits of threats, or pulling strands of analysis together into a particular whole. Such analysis requires a much greater staffing complement that currently exists. By way of comparison, I would point the Subcommittee to the National Security Council staff, which contains a series of well-staffed directorates for monitoring international threats, as well as a well-developed administrative infrastructure. Although it is fair to argue that it has taken more than a half century to develop, the nation does not have another fifty years to wait. This problem is particularly significant given the Office's role in maintaining the Homeland Security Advisory System, which assigns threats on a green (low) to red (highest) level.
6. Dissemination. The Office of Homeland Security appears to have adequate capability for disseminating information about potential threats.
7. Classification and Declassification. Although the Office of Homeland Security has ample authority to classify information as "top secret," it does not have parallel authority to order the immediate declassification of information that it deems in the national interest.
Recommendations for Action
I do not believe statutory authority can solve all of these problems. However, it can improve the odds that the Office of Homeland Security can produce the right information at the right time. At a minimum, I strongly recommend that the Office be given the authority, and budget, needed to produce and procure its own analysis. Some may argue that such analysis would merely duplicate already existing information. I would argue to the contrary. The Office of Homeland Security has a special obligation to examine information through a very broad lens and from a vantage point that no other agency of government has.
I also strongly recommend that the Office be given enough staff to interpret, analyze, and assay the information that it currently receives, and to make recommendations to Congress and the executive regarding improvements in the information system. That might mean, for example, that Congress would provide the funding for a minimum staff floor, or create a special information directorate within a new Office of Homeland Security. That might also mean that Congress would require the Office to conduct an information audit of the federal agencies it coordinates, and make recommendations regarding the technology investments needed to assure secure, high-quality information. Finally, I recommend that the director of Homeland Security be given limited authority to declassify information deemed in the public's interest.
These authorities will be wasted, however, if the current Office or its statutory successor is allowed to thicken with needless bureaucracy. The thickening has already begun. Being lean and flat is not just a value that the president himself espoused in the 2000 presidential campaign; it is also the sine qua non for effective information sharing. For whatever reason, the current Office of Homeland Security has already become one of the thickest units in the White House. That not only weakens communication within the Office, it merely adds to the extraordinary thickening of other units of government. To the extent the president relies on Gov. Ridge to stay in touch with the front-lines at the INS, for example, he is staring down an information chain with 30-35 links. That is more than the nation can afford. APPENDIX
At least in organizational terms, the department has long subscribed to the notion that wider and taller is better. At its founding in 1979, its secretary, deputy secretary, under secretary, and assistant secretary compartments contained ten layers, and 56 senior executives, both political and career. By the winter of 1998, those four compartments had thickened to 18 layers, and 143 senior executives, including an assortment of chiefs of staff and other alter-ego deputies who fill in whenever their principals are out.
The diffusion of accountability hardly ends at the government's door, however. It continues down into the University of California, the contractor that supervises the Los Alamos laboratory and three other DOE facilities. Who does the federal government hold accountable at the university? The president? The senior vice president for business and finance? Vice president for financial management? Associate vice president for human resources and benefits? Assistant vice president for laboratory administration? The executive director for laboratory operations? Director of contracts management? Manager for facilities management and safeguards and security? And so the questions mount as one works down the seemingly endless list of divisions, bureaus, and offices responsible for overseeing the laboratory and its collection of secrets.
No wonder it takes a crisis to focus attention. With 25 layers just to get from the top of the department to the top of Los Alamos, information is bound to get lost along the way. Like the childhood game of telephone, in which messages become hopelessly distorted as they are relayed from child to child, the layers merely add to the potential confusion and loss of accountability.
The Department of Energy is hardly alone in the senior-level thickening, however. Forced by repeated hiring freezes to choose between protecting the bottom of government or bulking up its middle and top, federal departments and agencies have mostly sacrificed the bottom. Nineteen-ninety-seven was the first year in civil service history where middle level employees actually outnumbered bottom-level employees. That giant sucking sound coming from Washington over the past few years has been the retirement of nearly 200,000 senior and middle-level managers government and the promotion of almost everyone next in line, all at a cost of $3 billion in voluntary buyouts for what turned out to be a big retirement party with no visible impact on the basic structure of government.
Some of the lower-level jobs have disappeared forever with the arrival of time-saving technologies. Others have migrated upward into the middle-level ranks where professional and technical employees have added lower-level tasks to their higher-paid duties. Still others have migrated outward and downward into the federal government's contract workforce, which numbered an estimated 5.6 million employees when President Clinton appeared before Congress to declare the era of big government over in 1996.
Meanwhile, the top of government has grown ever taller. From 1993 to 1998, federal departments created sixteen new senior-level titles, including a stunning number of new "alter ego" deputy posts, including "deputy to the deputy secretary," "principal assistant deputy under secretary," "associate deputy assistant secretary," "chief of staff to the under secretary," "assistant chief of staff to the administrator," and "chief of staff to the assistant administrator." All totaled, the administration created as many new titles in its first five years as the previous seven administrations created in their thirty-three years combined.
Reacting to news that his Immigration and Naturalization Service had belatedly mailed student visas to the flight school that trained two of the New York City terrorists, Commissioner James W. Ziglar immediately transferred four career employees involved in the snafu, and promised to fix the agency once and for all. "The breakdown in communication, highlighted by this week's events at INS, is unacceptable and will not be allowed," he said.
If Commissioner Ziglar really wants to know what went wrong in the incident, he merely needs to open the INS phone book and start dialing. He should start by interrogating his own chief of staff, his director of the executive secretariate, the deputy commissioner's office, and the counselor to the deputy commissioner.
He must have talked with his executive associate commissioner for field operations before firing him last week, but should still call his deputy executive associate commissioner for investigations, assistant commissioner for intelligence, executive associate commissioner for management, and the long list of associate commissioners, assistant commissioners, and deputy assistant commissioners that help manage the agency.
Ziglar could always call upward to the attorney general for advice, but that would require dozens of phone-calls through the thicket of political and career executives interposed between INS and the top of the Justice Department. Ziglar would have to get through the associate attorney general, eight associate deputy attorneys general, one principle deputy attorney general, the chief of staff to the deputy attorney general, and the deputy attorney general himself before working past the deputy chief of staff, chief of staff, and assistant to the attorney general himself.
The more Ziglar lets his fingers do the walking, the more layers of management he will uncover. He is at least six layers from the top of his department and twelve from the bottom of his agency. Back in 1995 when I counted every last layer in the agency, policy direction passed through 65 different people on the way down to each immigration inspector, while budget requests passed up through 62. And that was before the agency created a chief of staff to the commissioner. In such an over-layered hierarchy, Ziglar could not prevent a communications breakdown if he wanted to. The odds are that he had never heard of ACS, Inc., until last week, and that ACS knew more about the Kentucky congressional delegation, which took justifiable pride in the INS contract, than Ziglar.
Ziglar's problem is not unfamiliar in government. It was the central cause of the taxpayer abuse at the Internal Revenue Service in 1997, the security breakdowns at the Department of Energy in 1999, and the lack of communication among a host of agencies before September 11.
Ziglar and Congress cannot solve this problem one agency at time. Rather, they need to take an ax to the entire super-structure of government. President Bush started the effort last spring when he ordered every federal agency to prepare a "restructuring plan" designed to "flatten the federal hierarchy, increase the span of control, reduce the time it takes to make decisions and reduce the layers in government."
The order was mostly forgotten after September 11. Moreover, by ignoring political appointees, Bush missed at least a third of the bloat that interferes with citizen service. Political appointees may account for a tiny fraction of the total federal workforce, but they occupy between a quarter and two-fifths of the bureaucratic layers in the individual agencies of government, including 41 percent of the bureaucratic layers between president and the front-lines at INS. Either layers are a problem or they are not.
It is not too late to rescue the Bush effort, however. September 11 creates an obvious urgency for compliance, and the president's homeland security chief, Tom Ridge, has the kind of determination to make such the flattening stick. Ridge is supposedly working on reorganization proposals to create a new border security agency composed of INS, the Border Patrol, Customs Service, and Coast Guard. It hardly makes sense to bring those agencies together without a commensurate flattening both inside and outside the new agency. At least for now, the fog of war that confuses the troops is not to be found on the battlefields of Afghanistan, but in the bloated hierarchies of INS and its sister agencies.