Building and transferring expert knowledge within federal agencies

“If the IRS started staffing up now, it would take them at least a decade to catch up. They don’t have enough IRS agents with enough knowledge to know what they are looking at. They are so grossly overmatched it’s not funny.”  If this comment from Monte Jackel, former Special Counsel for the Office of Chief Counsel at the IRS, is not disconcerting, consider also the billions of tax dollars that are uncollected as the IRS performs fewer and less effective tax examinations.

The IRS is not the only federal agency dealing with the tide of employee attrition and the loss of vital institutional knowledge essential for advancing agencies’ missions. We use the IRS as an example so that we may suggest some ways that knowledge loss can be addressed. But the issues discussed below apply to any organization that loses experienced personnel.

Unfortunately, the IRS, like other federal agencies, cannot simply hire their way out of the problem of obtaining the talent and knowledge necessary to address complex mission imperatives. Certainly, new hires bring valuable skills, including knowledge of forensic accounting and emerging technologies such as A.I. and machine learning. Nevertheless, new hires do not possess the intricate skills and innate sense developed over years of experience by senior revenue agents in conducting complex tax examinations.

The senior revenue agent in performing a tax examination can expeditiously and accurately recognize what salient information is there, what isn’t there, and what should be there based upon context in the specific taxpayer’s industry. By recognizing situations where the facts presented by the taxpayer do not make sense, are not consistent with industry standards, or are contradicted by other evidence, the senior revenue agent is not led astray by irrelevant issues that sophisticated tax filers often use as smoke screens. Thus, Information Disclosure Requests formulated by such senior revenue agents save time and effort by focusing on truly germane tax issues. The senior revenue agent also knows when and how to request and deploy technical assistance such as computer forensic teams, valuation experts, and international examiners, and not waste time and effort attempting to single-handedly tackle a problem that may require the input from multiple technical disciplines.

Therefore, talented newcomers need help obtaining the kind of know-how that senior revenue agents have accrued. There are several challenges to developing that kind of knowledge.

A major challenge is that while a great many procedures and guides to identify and address specific issues have been written down, much of what highly experienced revenue agents know has never been documented and resides solely in their heads—including when and exactly how to apply rules and procedures. New hires and junior revenue agents who are less skilful in the performance of tax examinations or other services simply do the best they can. But that best can result in mistakes of omission or commission.

In an ideal world, senior revenue agents would guide long-term apprenticeship programs to pass along their institutional knowledge. For example, the IRS would benefit from utilizing senior revenue agents to serve as coaches by embedding them in work units to guide junior revenue agents in conducting examinations and making recommendations to improve journeymen efficacy and skill in conducting successful examinations. The IRS did successfully utilize this approach in the late 1980s and early 1990s through the Coordinated Agent Position (CAP) Program, in which senior revenue agents were assigned to work with existing IRS work units. The CAP Program was ultimately discontinued because IRS senior leadership perceived that it required taking senior revenue agents away from performing their own examinations and achieving tax adjustments in major cases. A possible solution would be greater reliance on reemployed annuitants to coach junior revenue agents in conducting examinations.

If IRS leadership determines that it cannot afford to dedicate the time of senior revenue agents to help develop junior revenue agents, we suggest alternative approaches that have been implemented at many different organizations. Given that so much essential expertise is undocumented, there is simply no way to completely avoid engaging the key talent in sharing their knowledge. However, a transfer process can be designed to minimize the expert’s time and place more responsibility for knowledge transfer on less experienced colleagues.

For example, one company that we will call Alpha Engineering and Architecture (“Alpha”), faced a similar challenge to what the IRS confronts: the loss of top talent to retirement and the need to capture the key employees’ critical business knowledge. However, these individuals were extremely busy carrying on the work of the firm and couldn’t commit to more than a few hours every two weeks for mentoring.

In such situations, focus and structure are key. Not all the expert’s knowledge can or should be transferred. Rather, specific experience-based skills are targeted—ones involving judgment, diagnostic capability, and ability to anticipate downstream implications. Once the targets are established, the interactions between experts and those who wish to learn the expertise are carefully structured to make the time as productive as possible.

At Alpha, junior engineers and architects were paired with an expert. The two engaged in an accelerated apprenticeship. First, field-tested interview protocols identified and prioritized the experts’ key areas of knowledge, selected based on criteria such as criticality to operations, degree to which these areas could be addressed by hiring from outside the company and imminence of expert departure. The expert/learner pair then designed specific mini-experiences for the less experienced employees, in order to elicit and share the critical know-how within the selected knowledge targets. These mini-experiences were often built into normal operations. For example, one learner began attending all meetings with her designated expert to learn the essentials about an important client. The pair planned brief internal follow up meetings to discuss the nuances of how the expert handled the relationship and how to avoid certain pitfalls. The limitation of such an approach is that it cannot fully convey the depth of knowledge and insights possessed by an expert that would be possible in a more extensive coaching context such as the CAP Program. However, the tight focus on just a few key aspects of expertise at a time and the structured process limits the time that experts are drawn away from their work.

There are other techniques for capturing and diffusing expertise that lessen the time burden on experts. One is to create a “knowledge cascade.” The knowledge starts its journey through the organization when a designated learner analyses interview recordings (or notes) of relatively short, carefully designed interactions with a single expert in order to extract rules of thumb, guidelines, checklists, process steps or cautionary tales that reflect the expert’s experience. The knowledge thus captured is passed along in a “pay it forward” model, as the initial recipient teaches others and/or posts wikis and blogs about the key points learned.

At Alpha, for example, a learner organized a series of “campfire learning” sessions. He asked several experts to talk with a group of junior engineers on a specific topic for an hour. Each session was recorded and uploaded to the company intranet open to anyone interested in the topic. The learner also analyzed the campfire discussions to create guidelines or checklists never before elaborated by the expert. One such campfire focused on how to conduct walk-throughs of existing hospital facilities in order to design renovations for a medical facilities client. The expert enumerated a wide variety of existing conditions that must be noted and addressed in the renovations, such as traffic flow, nurses’ access to supplies, accommodations for patients’ families, and patient privacy. While the expert had a well-honed mental checklist and could conduct a walk-through at a speed and degree of thoroughness unattainable by her colleagues, no one had ever captured this know-how until the campfire convenor created a written checklist. At Alpha, such checklists became the focus for discussion and further refinement by junior employees. This mode of diffusion amplifies every expert’s otherwise unshared wisdom and ensures that experience-based knowledge is not lost when the expert exits the organization.

An investment in knowledge transfer recognizes the value of institutional knowledge and the dedicated public servants who possess this knowledge. Federal agencies like the IRS cannot solely rely on hiring or web-based distance learning to build capability when the success of occupational roles such as revenue agents depends so much on the use of undocumented knowledge. IRS revenue agents, like other career public servants, are not fungible; and their experience-based knowledge cannot be taken for granted. Knowledge transfer is a vital organizational imperative that the IRS and other federal agencies should recognize and implement.

Dorothy Leonard is the William J. Abernathy Professor of Business Emerita at Harvard Business School and co-founder of the Leonard Barton Group consulting firm focused on improving knowledge management and innovation within organizations. 

John Malgeri, J.D., Ph.D., is the Co-Chair of the HRStat Community of Practice. 

Cayly Dixon, MS, is an Associate of the Leonard-Barton Group and founder of Conoc Partners, a knowledge transfer consulting firm.

Comments