Hermann Habermann

Positioned in the White House Office of Management and Budget (OMB), the CSOTUS plays a critical role for the US decentralized statistical system—13 principal federal statistical agencies and scores of statistical units across the federal government—ensuring the system operates smoothly and efficiently, is seen as a coherent unit with a strong voice, and takes advantage of synergies to address shared challenges. Foremost among the chief statistician’s congressionally mandated responsibilities is overseeing the production and accessibility of objective, reliable, and timely statistics.

Katherine Wallman

The CSOTUS also promulgates governmentwide standards that foster comparability and utility of the data, coordinates US participation in international statistical activities, and reviews and approves statistical agency information collection requests.

As of the date of this publication, 27 months since Potok’s retirement from the post, no one has been named as CSOTUS.

Nancy Potok

The challenges the CSOTUS faces are high in the minds of Hermann Habermann, Nancy Potok, and Katherine Wallman. The three think the organizational positioning of the CSOTUS within OMB, schematically shown in Figure 1, does not adequately provide the holder of the position the access or leverage needed to exert its many authorities provided by law. They recommend strengthening the position and discuss two options. They also note how strengthening the CSOTUS position is especially needed now, with the recent creation of new roles and offices in OMB on evidence, information, and data and the proliferation of private sector and intergovernmental data sources and providers.

The chief statistician’s small staff size is another major challenge for fulfilling the maximum potential of OMB in evidence-based policymaking across the federal government and beyond, including collaborating with state and local governments and the private sector.

To aid the understanding of the current challenges, Wallman assembled a brief historical timeline of the position, which is shown in the sidebar of this article. With its responsibilities originating in the 1930s, the chief statistician steadily accumulated responsibilities, most recently with the Evidence Act of 2018. Its organizational positioning has bounced around a bit, including a brief stint in the Department of Commerce, before settling in its current umbrella organization, the OMB Office of Information and Regulatory Affairs.

When asked about the state of the CSOTUS, Habermann, Potok, and Wallman preferred not to answer the question at this point, noting the vacancy in the position for more than two years indicates it has not been a priority for either the prior or current leadership at OMB. They did note, however, there is always the potential for change when a new CSOTUS is installed, particularly if the position can be strengthened and adequate staff is provided to carry out the statutory responsibilities of the job.

For the next CSOTUS, Habermann, Potok, and Wallman offered sage advice for how to be effective in the role.

The State of the Data Infrastructure Series frames the federal statistical agencies as the backbone of the US data infrastructure. Just as our transportation infrastructure supports the US economy, governance, and society, so too do the federal statistical agencies. Other installments can be found on the ASA website under the bar labeled “What the statistical experts say.”

Please describe the most prominent or important roles of the chief statistician and why they are important.

Katherine Wallman: First among the responsibilities of the chief statistician is overseeing the production and accessibility of objective, reliable, and timely statistics that garner the trust of our nation to inform discussions ranging from health and economy to education and security.

While much of the collection, processing, and dissemination of statistics produced by the federal government takes place in the dispersed US statistical agencies, establishing priorities and securing budgets for these activities rest in large measure on the chief statistician. Likewise, it is the responsibility of OMB and the chief statistician to promulgate government-wide standards that foster comparability and utility of the data. These range from standards that safeguard the autonomy and objectivity of statistics to those that facilitate the use of common classifications to others that ensure equitable access to information.

Hermann Habermann: As a believer that official statistics are critical to the lifeblood of democracy and because we have a decentralized statistical system, the chief statistician provides both the vision and the leadership to that system to ensure a healthy federal statistical system. This to me is the most important function.

Nancy Potok: I agree with Katherine and Hermann and would take it a step further: The chief statistician is also a bridge builder to the data communities outside the statistical system. In today’s information and evidence-building world, many statistical and information-related activities are taking place outside the federal statistical agencies. Many of these players would benefit from the decades of experience the statistical agencies have acquired and learned on such topics as standards, quality, ethics, confidentiality, and applications of data. So, in addition to looking inward at the federal statistical system, the chief statistician should also be externally facing and making sure the federal statistical viewpoint is fully represented in the evidence building and data science ecosystem.

Conversely, the chief statistician also needs to be doing everything possible to provide the resources and enable an environment in which the statistical agencies move with alacrity to keep up with advances being made outside the system, stay relevant, and meet the expectations of policymakers.

Hermann Habermann: That’s an important point. Thirty years ago, the federal government controlled most of the information about people. Now, the private sector controls it and dominates where the nation is going with respect to privacy and technology. I agree with Nancy about the need for the chief statistician to provide the vision for the federal statistical system and to be a major voice in the national debate about the use of official statistics, including privacy and confidentiality of individual information.

Please share an example of the key role of the chief statistician.

Katherine Wallman: Parallel with the growth of technology and data sources came the establishment of chief data officer, chief information officer, and chief technology officer roles. There was also the generally sensible movement to centralize IT across a department for efficiency and cost savings. Such centralization, however, is a concern for federal statistical agencies, which store data that has been collected under pledges of confidentiality. When the National Center for Health Statistics (NCHS) was wrapped into the IT centralization efforts of the Department of Health and Human Services (DHHS), I worked alongside the NCHS to maintain the reality and perception of the care that’s taken with the data the public provides in the particularly sensitive area of health. We fostered what we referred to as a “two-key solution”; IT people in DHHS couldn’t get access—albeit only for system maintenance—without having someone from NCHS present. Our solution also opened the door for similar approaches in other statistical agencies faced with centralization of their departments’ IT systems.

Hermann Habermann: An example I hope helped not just one agency, but the entire federal statistical system, is the creation of the Joint Program in Survey Methodology. While the academic statistical community in the United States is rich, several of us thought there was a need for a program focused on survey methodology for official statistics when I was chief statistician. OMB, due to its unique position in the executive branch, afforded us the opportunity to create this program.

Nancy Potok: In the work to legislate the recommendations of the Commission on Evidence-Based Policymaking in the Foundations of Evidence-Based Policymaking Act, evaluation and building evaluation capacity for evidence building were an early focus. So, during the drafting of the legislation, we were able to achieve three important advances for federal statistics. The enacted bill, now generally referred to as the Evidence Act, put into statute a key section of OMB Statistical Policy Directive #1: the need for independence of the statistical agencies for them to ensure the objectivity and integrity of their products.

In addition, to complement the creation of chief data officers and chief evaluation officers, the Evidence Act created a senior statistical official in all the cabinet agencies, even those without a designated statistical agency.

Finally, the bill really strengthened and updated the Confidential Information Protection and Statistical Efficiency Act of 2002 (CIPSEA), a crucial element strengthening the federal statistical agencies that was first enacted under Katherine’s watch. The 2018 reinforcement and update, along with additional key roles for the statistical agencies in evidence building, is crucial to enabling the federal statistical system to modernize if the bill’s provisions are implemented in an inclusive way for the wider evidence-building community. That wider implementation would benefit the public good with a synergy that is hard to achieve in our decentralized and fragmented system. This statutory strengthening of the federal statistical system probably would not have been achieved without the chief statistician in OMB being at the table, as Katherine puts it, for those discussions on the bill.

What are the primary challenges to the position of chief statistician and staff?

Katherine Wallman: I will start where Nancy left off. A fundamental challenge for the success of any chief statistician is getting a seat at the table, especially in the emerging forums Nancy mentioned earlier. To be effective, the chief statistician has to get into the right rooms and conversations and cochair the right groups. Among other things, such positioning takes support from OMB and the community, but personality and persona also go a long way.

I also think the chief statistician is challenged to take advantage of the wealth of expertise across the federal statistical system. Based on my experience sitting at the agency heads’ table, the heads understandably prioritized the work in their respective host departments. Whatever they did on an interagency basis was beyond their “regular duties.”

I believe responsibilities for interagency work should be part and parcel of agencies’ personnel portfolios. In our decentralized system, we have to get that valuable expertise into the room. With the chief statistician’s relatively small staff, that expertise, frankly, is often greater in the agencies than in the staff at OMB. There are advantages to using that expertise that go beyond “just” the person power, but I’m not sure that’s perceived equally by the agencies providing that expertise.

Hermann Habermann: With one major exception, I think the Chief Statistician of the United States has a powerful set of tools. If the chief statistician is able to leverage the development of the federal budget and the Paperwork Reduction Act (PRA) process, they have an incredible amount of influence. Let me give an example of the importance of being a part of OMB.

When I created the Interagency Council on Statistical Policy (ICSP), everybody I invited came. This was partly because it was OMB doing the invitation. The heads of the statistical agencies also recognized the good that could be accomplished through creation of the ICSP. What I think is missing is that the chief statistician—a civil service position—doesn’t have a recognized place at the political table, despite the position’s responsibility both internationally and nationally.

OMB policy decisions are made at the level of the director and associate director. If the chief statistician were a (term-limited) political position, it would enable them to be a member of the policy team within OMB and provide greater access to other parts of the Executive Office of the President and the White House.

Katherine and Nancy did a much better job than I trying to get a place at the table. But I think they were hindered because the position is civil service. And no matter how much you try, if you’re not a political appointee, it’s not the same thing. That’s a big hindrance. I believe the chief statistician should be a political appointee with a fixed term to have a political place at the table. That to me is the biggest challenge.

Nancy Potok: I agree the chief statistician potentially has many powerful tools. I also agree that where the chief statistician is situated is a hindrance to maximizing the use of those tools and the chief statistician is often shut out of policymaking due to being a civil servant.

As for whether the chief statistician position should be a political appointment, I think there’s a great danger in that because political appointees in OMB are taking their orders directly from the president. And presidents have demonstrated they are susceptible to the temptation of skewing statistics to support their political agendas. If the chief statistician could not hold that line on independence and objectivity, the credibility of US statistics would be at stake. This would have far-reaching repercussions on international and domestic investment in the US, the stock market, and a whole range of decisions being made that depend on federal statistics being considered the “gold standard.”

We’ve seen other countries where official statistics can be manipulated with bad intent, and it leads to corrupt governments and lack of credibility on the world stage. The challenge to me is finding ways to ensure you get somebody who is not going to have political motivation into a political appointment. Even if that can be achieved four out of five times, you could have a real problem when you get the fifth. The whole system could be dismantled and discredited with an administration that doesn’t want to observe norms. So, you would have to have some statutory protections, such as strong qualifications, a fixed term that crosses election dates, and so forth. Maybe a Senate-confirmed, nonrenewable 10-year term like that of the FBI director would provide a measure of independence where needed while still respecting the power of the president over the executive branch.

Elevating the career chief statistician position within OMB is another possibility. I point to the September 2020 report of the National Academy of Public Administration task force, which consisted of bipartisan former OMB political appointees and career executives who worked in different functional areas at OMB. I cochaired the task force with Nick Hart and we came to a consensus that OMB’s information policy work should be restructured and reprioritized by bringing together statistics, evidence building, and IT under a new senior position we called assistant director for information policy. We also concluded the position of chief statistician should be elevated from a career branch chief position wherever it landed in OMB. The chief statistician could easily be dual-hatted as the assistant director for information policy.

The task force was formed out of a recognition for how much has changed since the enactment of the Paperwork Reduction Act in 1995, including the creation of chief data officers, chief information officers, and an evidence team in OMB. Putting these functions under one office minimizes the potential for centers of fragmented power in OMB that, in turn, lead to potentially confusing policies and directions for agencies that are hard to tie to budget priorities. The new position, like the current assistant director for budget, would be a senior career person with the influence to attend political appointee–level meetings, thereby representing and injecting federal statistical information and evidence into public policy decisions and helping represent an intergovernmental perspective on data collection and use. The chief statistician could either be the assistant director position or report directly to that position.

What’s most important is elevating the status of the chief statistician. The current location in the OMB management hierarchy is inadequate in terms of achieving that seat at the table and the influence needed to be effective for all its many roles. The new positioning within information policy would be a higher-level, more visible position for the chief statistician. Although a restructuring within OMB would mean moving the chief statistician out of the Office of Information and Regulatory Affairs, some thought would need to be given as to how the transactional aspects of PRA, especially the reviews of information collections, would be handled and coordinated.

Another challenge is there’s not nearly enough staff to do the work. The competition for resources in OMB is intense, but the chief statistician’s current staff size is having a negative effect on the ability to accomplish important work. For example, regulations required by the Evidence Act are years late in being issued. Having the position in an elevated information function may help with gaining the staff needed to meet the statutory responsibilities of the job.

Katherine Wallman: On Nancy’s point for how much OMB has changed over the decades, I’ll note that—before the PRA—the statistical policy office had two broad responsibilities: developing and promulgating statistical standards and reviewing all information collections, the latter of which started with the 1942 Federal Reports Act.

When the regulatory review function was created and assigned to OMB, it was all consuming. Due to artifacts of history that took place in the late 70s, regulatory review was coupled with statistical policy—a rather clunky placement in my view. I agree it was detrimental to statistical policy work and that separating them would be advantageous for statistical and information policy.

With respect to staffing in the office of the chief statistician, the small number made it all the more critical to have people at the top of the civil service who had a lot of experience in statistical agencies in particular. With such backgrounds, statistical policy staff members were both knowledgeable and generally known across the system and thus could accomplish the responsibilities of the office effectively. Over the years, because such hiring practices contrasted with those in other parts of OMB, justifying replacement hires at relatively senior levels could be a challenge.

What are the consequences of these challenges not being addressed (both in terms of what couldn’t be done in the past and what potentially will not be prioritized going forward)?

Nancy Potok: Many opportunities are being missed. I already mentioned the ability of the chief statistician to be a bridge to the broader data community. I would also point to holding together the disparate efforts across the federal government that are data science–oriented and need input from the statistical community to optimize how they’re being researched, developed, and implemented across agencies. The federal government tends to have a siloed viewpoint. For example, for IT modernization several years ago, it was all about hardware and software and missed the mark when it came to thinking about the content of the data residing on that infrastructure.

The second area is intergovernmental relations, because OMB traditionally doesn’t have the functions that reach out to state and local governments. But through a data lens, some of the most valuable data resides in state and local government. Federal data collections also have their problems in that they overlap, are redundant, and don’t really collect the most meaningful data (e.g., programmatic data from state and local agencies that are the beneficiaries of federal funds). Such opportunities are all the more valuable with administrative program data becoming more important in creating statistical data sets. Currently, there’s no ability for the chief statistician to influence those in a meaningful way.

Thinking about the data landscape as an ecosystem is important as one considers the many parties creating, disseminating, and using statistical data. Besides technology, there are data collections coming from state and local governments that are beneficiaries of federal funding and administrative program data coming from individuals, such as tax and social security data. The private sector has amassed huge amounts of data on both individuals and businesses. All those factors need to be thought about more holistically to do a better job of creating relevant, timely, objective, high-quality, and ethically sourced and used statistical data. That’s an opportunity we’re losing out on by limiting the utility of the chief statistician position.

Hermann Habermann: I agree with Nancy that we’re missing opportunities. For example, many universities are moving away from just statistics and toward the integration of data science and statistics. The traditional role of the federal statistical system in being the sole or primary producer of the nation’s seminal statistics is being challenged. The private sector is not only producing data that was once the sole province of the federal statistical system, but it is offering timely and small area data in many cases.

Along with much more private sector data, many more people are analyzing data and producing statistics. OMB leadership is needed, for example, to understand what partnership with the private sector means and to what extent federal agencies should also become brokers—to inform the political leadership about how to interpret the data coming out of the private sector. These are the kinds of questions that are not being addressed by not having a chief statistician at OMB for two years now, resulting in a vacuum at OMB in terms of addressing solutions to these kinds of problems.

Katherine Wallman: I agree with Nancy and Hermann—particularly as they have noted the silos that unfortunately seem to increasingly characterize the federal roles in statistics, data, and information policy and the crying need for effective partnerships with other levels of government and the private sector.

As a specific example of the seeming failure to coordinate among federal agencies, a recent announcement called for comments concerning an agency’s desire to do data linkage between data sources. It wasn’t clear to me to what extent the requesting agency was aware of and connected to the extensive data-linking efforts underway by the federal statistical agencies. The chief statistician frequently plays a role in fostering communication so agencies aren’t reinventing the wheel and can learn and benefit from one another’s work—but many opportunities to facilitate communication and collaboration are being foregone. Similarly, work to measure and standardize the quality of combined data isn’t being pursued with as much speed and energy as it would with a chief statistician in place.

A new CSOTUS may soon be named. What recommendations do you have for this person for carrying out their responsibilities, given the state of the CSOTUS position?

Nancy Potok: First, educate yourself about how we got to where we are today and understand the evolution of the system. Learn about why the standards are there, not just that they exist. Learn about and understand the landscape in terms of what’s available to you to leverage. Then, set your priorities and stick to them. It’s easy at OMB to react to the crisis of the day and lose the focus needed to accomplish your own goals.

Second, rather than acting like an OMB branch chief, which doesn’t have much say in OMB, act like you’re the Chief Statistician of the United States. That is, take advantage of your significant and important external presence. You can accomplish a lot if you don’t limit yourself—put yourself at the table to the extent that you can. Even if you aren’t invited to the meeting, find a way to make sure you’re there. Assert your authority as the chief statistician.

There were a lot of people I met who were new appointees in the administration. I just said, “I’m the Chief Statistician of the United States, and I started in January 2017.” They assumed I was a political appointee and acted accordingly, and I never corrected them.

Third, make sure you keep all your channels of external communication open with the stakeholders in the broader community—including local government, state government, and data users—and use them extensively. Then, of course, take seriously what people are telling you outside the system. Build relationships with your colleagues in other parts of the Executive Office of the President, such as the Office of Science and Technology Policy and Domestic Policy Council. Finally, to the extent you can, keep your lines open with Congress either directly or indirectly.

Hermann Habermann: I think that’s very good advice, Nancy.

Katherine Wallman: I’d subscribe to your recommendations wholeheartedly, Nancy. There are challenges, to be sure. But there are boundless opportunities. Taking us back for a brief reprise, there are many authorities and responsibilities that can fully command the chief statistician’s attention; keep the vision.

Stakeholder Perspective: Nick Hart

We invited Nick Hart, president of the Data Foundation and a leading champion in the evidence-based policymaking community, to give the stakeholder perspective for this month’s State of the Data Infrastructure Q&A. He was previously director of the Bipartisan Policy Center’s Evidence Project and the policy and research director for the US Commission on Evidence-Based Policymaking.

Nick Hart

The role of the Chief Statistician of the United States has grown substantially in significance over just the last few years and is one of just a handful of positions at the White House Office of Management and Budget (OMB) explicitly established in federal law. The leaders who held this role over the past several decades set the stage for the position to serve as a steward of the statistical system. Congress codified this role with specific duties that were strengthened and expanded in 2019 with the passage of the Foundations of Evidence-Based Policymaking Act, or the Evidence Act.

Today, it is the responsibility of the chief statistician to be a key leader for the federal government and the country—not just the statistical agencies—in promoting effective implementation of that law. The chief statistician also has a responsibility to support evidence building for the whole of government, with statistical agencies providing a critical function and infrastructure that bolsters the capabilities for program evaluation, analysis, research performance management, and much more. It is also no accident that, in October 2021, the Federal Advisory Committee on Data for Evidence Building recognized the vitality of the chief statistician and even recommended that OMB elevate the level of seniority of the position, given the extensive duties and responsibilities in supporting evidence-building activities in government.

Clearly the amount and type of support staff at OMB for the chief statistician is a major limitation and should be reviewed by the OMB director and Congress, particularly given the substantial new responsibilities imposed by the Evidence Act. The fact that such incredible work has been accomplished over the years from such a small staff given such tremendous responsibilities is also a testament to the skill and expertise of those talented individuals. Limited capacity at OMB to support implementation of new expectations for information management, data collection, information quality, evidence-building activities, and other collaboration considerations must be addressed in coming years.

Simply put, the consequences of under-resourcing the chief statistician and federal statistical system cannot be overstated. For example, three years after enactment of the Evidence Act, OMB has not yet published draft regulations to enable the rules on public trust that Nancy Potok described [in the Q&A] or to enable new data-sharing authorities authorized in the Evidence Act for the statistical system. These important capabilities require expertise and staff at OMB to navigate the complexities of federal laws and regulations, the nuances of processes in statistical agencies and units, and realities of research activities in collaboration with experts across multiple disciplines.

Even more, the expert staff at OMB must work across agency silos and coordinate with emerging evidence-building leaders such as chief data officers and evaluation officers in formulating learning agendas and open data plans and publishing data inventories.

The gaps in staffing at OMB—and in the statistical agencies, themselves—can prove challenging when providing the leadership, coordination, and execution of these important tasks. This is not all to say that activities cannot be reprioritized in annual budgeting, but when it comes to the chief statistician’s office, it is clear more capacity is needed, given the growing system and demands placed on the infrastructure. A relatively small investment in these important priorities would pay dividends for the country’s ability to engage in improved data sharing, risk assessment, and sensitive data protection.

Allow me to add to the recommendations of Habermann, Potok, and Wallman for the newest chief statistician. When it comes to the Evidence Act’s success, collaboration is key. Reach out to the chief data officer and evaluation community to coordinate, help, seek feedback, and engage in dialogues about how to build a stronger evidence culture across government, leveraging joint resources and capabilities with other leaders.