NCES: Perspectives, Insights, and Recommendations from Former Leaders

Steve Pierson, Michelle Crosby, Stephen Provasnik, and Debra Silimeo

Amstat News published two articles in 2020 and 2021 about challenges the National Center for Education Statistics faced, one by ASA staff on bureaucratic and resource hurdles (“National Center for Education Statistics Faces Program Cuts”) and one interviewing three experts (“State of the Education Data Infrastructure: What Three Experts Have to Say About the National Center for Education Statistics”). The firing of NCES staff in March presents an opportunity to hear the perspective of recent top managers and leadership regarding challenges and potential solutions. We decided not to name those interviewed so they could speak candidly. The challenges NCES faced were long-standing, bureaucratic issues and not specific to an administration or individual(s), as noted in the introduction and documented in references in the sidebar. The perspectives of those interviewed provide additional context and insights.

In September 2025, ASA Science Policy Fellow Michelle Crosby guided a discussion with former NCES leaders who managed the agency’s operations through 2025. Over the course of the conversation, these leaders discussed various challenges they had encountered and built on one another’s observations. What emerged was a complex picture and concrete ideas for how a rebuilt agency could operate more efficiently. This article organizes their discussion into five topic areas: contracting and staffing challenges; redundant review; capacity and requirements mismatch; barriers to innovation; and IT challenges.

Contracting and Staffing Challenges

NCES struggled to function as effectively as it could because of unwieldy contracting and hiring processes imposed on it. This meant NCES could not efficiently contract for services or hire the staff needed to fulfill its legal mandates and the needs of data consumers.

NCES relied on contractors to conduct a significant amount of work because of the relatively low number of NCES staff relative to the volume of legal mandates.

Critically, however, NCES was required to work through the Department of Education Contracts and Acquisition Management Office for all its contracting activities and did not have a dedicated contracts office to support its needs. NCES and the Contracts and Acquisition Management Office often had conflicting and sometimes even contradictory goals. The Contracts and Acquisition Management Office saw its main mission as increasing competition for Department of Education contracts, which it believed was best accomplished by spreading tasks across multiple contracts, with a preference for fixed-price contracts, and helping new and small businesses secure contracts. NCES, in contrast, saw its main mission as ensuring the data they were legally required to collect was of high quality (and doing so with its limited staff size), which it believed was best accomplished by keeping down the number of contracts to manage, thereby bundling common tasks for efficiency, consistency, and cost savings.

These systemic conflicts manifested in a host of challenges for NCES. For example, NCES leadership noted the Contracts and Acquisition Management Office’s needs and guidance often increased NCES’s management workload and created significant interdependency challenges—such as creating situations where when one contract faced a delay, that delay would cascade to other contracts. The Department of Education’s contracts office also was short staffed and, at times, experienced significant turnover, resulting in contracting staff who did not yet understand NCES projects and procurement timelines. Though NCES would develop contract documents on a schedule to meet its data collection requirements, the Contracts and Acquisition Management Office was not reciprocally bound by NCES’s scheduling needs. These delays and challenges had real consequences with high opportunity costs.

Perhaps the most striking example of this involved a multi-year effort to create a multiple-award indefinite delivery/indefinite quantity contract vehicle—a pre-approved list of contractors that would allow NCES to award contracts more quickly. Development and award of the indefinite delivery/indefinite quantity involved a substantial number of NCES staff and went through the complete federal acquisition regulation process. The Contracts and Acquisition Management Office approved the vehicle and awarded the indefinite delivery/indefinite quantity contract after a demanding two-year effort, but then contracting personnel at the Contracts and Acquisition Management Office changed. A new contracting officer decided the indefinite delivery/indefinite quantity contract wasn’t acceptable for some types of work, even though the indefinite delivery/indefinite quantity contract intentionally included contractors who performed that work. As a result, for some new procurements, NCES staff had to go through time-consuming procurement steps that the indefinite delivery/indefinite quantity contract was intended to eliminate. For others, NCES staff had to go through all the procurement steps of the full contract award process for each new award (meaning all the work to create the indefinite delivery/indefinite quantity was for naught).

In another instance, the full field test for the 2020 cycle of the National Postsecondary Student Aid Study, which is legally required at least every four years, was scrapped because delays in the Contracts and Acquisition Management Office resulted in the contract being awarded too late for the field test to inform the main data collection. In a third example noted by NCES leaders, staff developed a proposal for a fixed-price-plus-incentive contract—which would reward contractors for exceeding performance metrics—but it was rejected because of the administrative burden it would create for Contracts and Acquisition Management Office staff.

More generally, Contracts and Acquisition Management Office turnover and short staffing made it challenging to set up new contracts or modify existing contracts to comply with new or changing congressional requests and address unanticipated challenges. Despite NCES having statutory authority to use alternative procurement routes, including the General Services Administration, requests to do so typically required approval from the contracts office and were usually declined. It was noted that NCES’s relationship with the Department of Education’s contracts office differed from other statistical agencies where, based on previous work experience, one leader described a more collaborative approach where program and contracting staff had shared accountability for adhering to timelines and awarding contracts that best met programmatic needs. NCES leaders also noted that having NCES’s staff hiring handled by the Department of Education’s Human Resources Office, which like Contracts and Acquisition Management Office supported NCES without a formal service agreement and upon which NCES was dependent without recourse, meant that errors and delays in the hiring process (completely outside of NCES’s purview) regularly caused NCES to go without key staff and to lose candidates to employers who could hire more quickly. For example, in 2024 NCES completed the months-long formal hiring process and approved a candidate for hire only to learn that the Department of Education’s human resources management canceled the hire because it believed its staff had made a clerical error in how it had classified and advertised the hire. In another case, also in 2024, a candidate NCES formally approved for hire was not notified of NCES’s decision by the Department of Education Human Resources Office for three months, by which time the candidate had taken another job.

Redundant Review

NCES strove to meet the demands and recommendations for timely information to make decisions. While progress was made, the review process imposed on NCES was a major barrier to the timely release of data products.

NCES already had a longstanding, rigorous, and comprehensive technical review process.

The Educational Sciences Reform Act of 2002 that created the Institute of Education Sciences specifies that NCES products must go through peer review, and this was operationalized as products being required to go through an additional layer of review in the Institute of Education Sciences Standards and Review Office.

The Standards and Review Office review added time to NCES’s release schedules because, in contrast to the review processes of the other 12 federal statistical agencies, it follows a journal-style peer review model, with reviewers often suggesting alternative analyses or re-running analyses with additional variables.

NCES leaders noted all NCES products sent to Standards and Review Office review had already been through extensive development checks, been reviewed by technical experts several times, and been checked by the NCES chief statistician.

Because NCES completes most of its work through contracts, Standards and Review Office–required revisions that call for additional academic work by the contractor was not always feasible given available contract funding, schedule, or both. In some cases, reviewer suggestions for revision were shelved, making the time allotted for Standards and Review Office review wasteful. In other cases, the product itself was shelved, even though NCES believed it was ready for release. Based on internal data, Standards and Review Office review added weeks or sometimes months to a product’s public release date. A former NCES leader explained this layer burns money and time and slows releases.

NCES staff sometimes chose product formats based not on what was best for a given topic or analysis but on what expedited the review process. For example, a tendency developed to use simpler release formats (e.g., data tables with no explanatory text other than table notes or two-page “Data Point” reports, rather than comprehensive analytical reports) because these formats faced fewer review hurdles and could speed up release dates by weeks or months.

Capacity and Requirements Mismatch

Despite being the third largest of the 13 statistical agencies by budget, NCES was ninth largest by staff size, which resulted in the responsibility for appropriated contract dollars being shouldered by fewer staff compared with other statistical agencies. The contrast is particularly evident in terms of a budget-to-staff ratio, which has been well documented by the ASA. Specifically, the FY23 size of NCES’s budget and staff translated into a budget-to-staff ratio of $3 million annually–eight times the median budget-to-staff ratio of $370,000 for the 13 statistical agencies.

Because of its high profile, the National Assessment of Educational Progress was better staffed. Its specialized organizational structure included dedicated staff for reading, math, science, scoring of each topic, administration, the testing platform, and sampling as well as multiple people with deep expertise in the topics necessary to support program operations. In addition, the National Assessment of Educational Progress internal administration was less challenging than for other NCES programs because the role of policy oversight, as outlined in the Education Sciences Reform Act, was assigned to the National Assessment Governing Board. This helped insulate the National Assessment of Educational Progress from the Institute of Education Sciences processes and procedures that often resulted in inefficiencies. As a result, the National Assessment of Educational Progress generally met its timeliness goals. The National Assessment of Educational Progress was also the only program in the assessment budget line, which was funded at $185 million in fiscal year 2025

In contrast, all NCES’s other survey work competed for funds out of the $122 million statistics budget line. The funding-level differences were reflected in staffing levels. For example, a complex study that included separate survey instruments for different types of respondents (e.g., public school teachers, private school teachers, and school administrators) and both national and state samples was handled by only two staff members.

Similarly, the longitudinal studies, which track students over time and involve multiple rounds of data collection, multiple components—parent surveys, school staff surveys, student assessments in various subjects—and development of reports and data products for every data collection, were overseen by, at most, two employees.

The technical review team that checked NCES products for quality faced similar constraints. A small group served the entire center and its approximately 60 data collection and reporting projects, which meant constant prioritization for review. The National Assessment of Educational Progress releases were always given priority. International studies were also prioritized when their internationally fixed release dates approached. For several months each year, technical reviewers focused heavily on materials for the Condition of Education, which met its mandatory June 1 release date every year until 2025. All other products had to wait for available capacity. NCES’s limited staffing precluded it having dedicated staff for the research and development that is so critical for a statistical agency to keep their work relevant and timely, a statutory requirement as well as an expectation of data users.

One reason for NCES’s severely limited staffing is that, unlike the other federal statistical agencies, NCES has not been allowed to use any of its funding to pay for staff. Since the establishment of the Institute of Education Sciences in 2002, NCES’s staffing level has been determined by the Institute of Education Sciences director, who has allotted NCES staff out of the overall Institute of Education Sciences staffing line set by Congress. Moreover, since NCES was placed within the Institute of Education Sciences, there has been no formal assessment to link NCES’s full-time equivalent needs with its workload.

Thin staffing meant NCES always seemed one retirement, one family emergency, or one unexpected technical crisis away from serious operational disruption.

Thin staffing meant NCES always seemed one retirement, one family emergency, or one unexpected technical crisis away from serious operational disruption. Critical institutional knowledge resided in a handful of long-tenured staff, and though NCES completed required succession planning, it lacked staff to implement the plan in a meaningful way. Moreover, survey operations had minimal surge capacity when problems emerged—meaning small issues could cascade into major delays.

Nevertheless, though staff-to-project levels were very low compared to other statistical agencies, NCES would not forego data quality controls needed to meet legal quality requirements and that are critical for maintaining the public’s trust in the accuracy of NCES information.

Institute of Education Sciences oversight often worked against the responsibilities of NCES to produce objective, timely, and trusted statistics in other ways. For example, NCES leaders said they wanted to make the NCES website more navigable and data easier to locate. But the new NCES website, launched in early 2025, was developed as part of an IES-wide website redesign that prioritized consistency across the four Institute of Education Sciences centers, in many cases making it more, rather than less, difficult to find NCES-specific tools and products.

Barriers to Innovation

NCES’s staffing and resource levels for the past two decades have left very limited, if any, capacity for research and development work such as testing new survey methodologies, improving systems, or experimenting with the adoption of AI outside of the National Assessment of Educational Progress program.

For example, NCES staff proposed to integrate the National Teacher and Principal Survey into regular state data collections of their public school personnel. Modeled on the efficiency and success of the Centers for Disease Control and Prevention program monitoring youth risk behavior, this proposal would have states collecting National Teacher and Principal Survey data as part of their normal operations, potentially making it more routine and consistent. This proposal could never move off the drawing board, however, without staff and funding for development work, exploring buy-in from states, and pilot testing.

Similarly, Congressional proposals to change the Integrated Postsecondary Education Data System from institution-level to student-level records collection could never be fully tested. Lacking staff and resources, NCES could not undertake the needed research and development to assess the feasibility of such a major change and inform design decisions across the entire collection lifecycle to ensure that the unit-record collection would yield the same high-quality data as the current collection, should Congress ever pass the legislation requiring a unit-record data collection.

For years, NCES leaders also wanted to incorporate more administrative data to reduce burden and costs, shorten release times, and improve data quality. However, because administrative data have their own challenges (for example, around issues such as how different administrators interpret instructions when reporting, lack of population coverage for groups not in a given program, and limited auxiliary variables), incorporating administrative data would have required focused R&D, particularly to gain a better understanding of how different methods of administrative data collection (e.g., state reporting versus “web scraping” or other data science applications) affect data quality and total survey error.

In each of these cases, NCES staff were qualified to undertake the needed research and development work, but they lacked the bandwidth to do so given existing duties. All staff were already fully committed to statutorily required work. NCES operational realities also impeded this kind of innovation: data collections could not be paused or significantly altered mid-cycle—stakeholders expected continuity. To conduct research and development, NCES needed the capacity to devote staff resources and funding for research and development without putting at risk its ongoing operations.

Information Technology Challenges

The requirement that NCES rely primarily on the Department of Education’s Office of the Chief Information Officer for IT services created additional obstacles to agility and product timeliness: requirements designed for policy and program offices didn’t necessarily fit statistical operations, approval processes stretched over years, and NCES was barred from developing its own IT-related technical capacity.

One extended discussion among former NCES leadership involved multi-factor authentication. To implement federal requirements related to this security measure, the Office of the Chief Information Officer indicated that all users accessing Department of Education IT systems needed to use multifactor authentication. Initial discussions focused on use of government-issued identification cards to authenticate users, which, for most Department of Education offices, made sense and was operationally feasible. However, for NCES, whose users included voluntary study participants such as fourth graders taking the National Assessment of Educational Progress and households randomly selected from an address-based frame, it was not. Not only was it not logistically feasible, but it raised concerns about the potential effect on already low response rates and legal requirements related to identification of participants in statistical data collections. Similar patterns emerged across other IT requirements: Department of Education policies appropriate for one context were applied broadly without accommodation for NCES’s different operational needs and legal requirements.

Software approval processes presented another challenge. One leader wanted to use Databricks, a tool for data processing that could significantly speed up work. The anticipated approval timeline: two years. The decision was made not to not to use Databricks, because it would take too long and the chance of approval too uncertain. The same constraints applied to artificial intelligence tools that could have improved efficiency, so NCES had to contract out capabilities they could have used directly.

When NCES tried to build internal IT capacity to work within these constraints, they also encountered barriers. NCES sought to hire staff with IT systems expertise to build and maintain data infrastructure, as well as to meet continually increasing requirements for IT security activities and documentation from the Office of the Chief Information Officer. However, NCES was told hiring for positions with such expertise was restricted to the Office of the Chief Information Officer—NCES could not hire people with those qualifications. This resulted in study staff, some without any IT expertise, having to add IT systems work to their list of responsibilities, and it left the agency dependent on a central office that did not always understand statistical agency requirements. Leaders noted that Federal Student Aid’s highly publicized IT implementation failures reflected broader challenges across the Department of Education that affected NCES, as well.

Recommendations

Based on these experiences, former NCES leaders identified specific changes that could make a rebuilt agency more agile and efficient. They noted that none of these ideas for changes are new; they had long discussed these issues but lacked resources or power, or both, to implement such changes.

Align staffing with program scope. Other federal statistical agencies, particularly the Bureau of Labor Statistics and Census Bureau, formally align their full-time equivalent with program requirements. NCES should be allowed to conduct a similar assessment, with transparent results that are used to inform staff allocations not only to traditional in-house functions but also to some previously done by contractors that could be more efficiently done in house.

Modify the review process. Leaders recommended either eliminating SRO review for statistical products or, if such additional review were needed, aligning it with the norms and standards of a federal statistical agency rather than those of an academic journal’s peer review. The leaders believe the legal basis for this already exists because the Institute of Education Sciences director has discretion in implementing the Education Sciences Reform Act’s peer review requirements.

Establish an R&D function. A rebuilt agency needs dedicated funding and staff for research and development—separate from the resources committed to collections and reporting that meet legal mandates. This capacity would enable testing of new approaches, administrative data integration research, survey redesign, process improvements, and methodology development. Innovation requires dedicated resources; it can’t happen in the margins.

Formalize service agreements. Whenever NCES needs to rely on shared services with other offices for IT, HR, contracting, and other support functions, it should have memorandums of understanding that document service levels, timelines, and escalation procedures. The memorandums of understandings should also specify when NCES can access alternative service providers. This would create clarity on responsibilities, provide a historical record for new staff, and establish frameworks for problem-solving. It would also make accountability clearer—if NCES is measured on timeliness, these enabling and support functions should have corresponding commitments to provide services in a timely manner. If NCES does not receive the quality of services it needs, it should be free to procure those services itself or from other agencies as outlined in the trust regulations.

Develop a comprehensive action plan. At the recommendation of a National Academies panel, NCES developed a five-year strategic plan but didn’t have the capacity to develop an action plan that aligns scope with resources, establishes clear priorities when trade-offs are needed, sets realistic timelines for major activities, and creates regular release schedules. NCES leaders recommend development of such a plan. Stakeholders would know what to expect and when. When resources don’t match statutory requirements, the plan would make the trade-offs transparent.

Use existing legal authorities to clarify the NCES-Institute of Education Sciences relationship. NCES leaders recommend that NCES be delegated the authorities required for it to efficiently carry out its work. As further discussed in ASA’s October comments to the department, the Evidence Act and Trust Regulation require parent agencies to “enable, support and facilitate” a statistical agency’s ability to meet its responsibilities, including delegation where needed. Delegation, NCES leaders explained, would make NCES a more agile agency better equipped to provide more timely products. They also recommended a memorandum of understanding between NCES, Institute of Education Sciences, and the Department of Education to delineate what is and is not being delegated, as well as ways that the Institute of Education Sciences and Department of Education will support NCES to fulfill its statutory requirement. Such a memorandum of understanding would include providing NCES its full statutory contracting authority, oversight of its information technology operations, and a budget proposal formulation process as is required in the evidence act.

Congressional guidance on priorities. The Education Sciences Reform Act, the now 23-year-old law, contains extensive “shall” statements requiring collection of data on various topics, plus a requirement to conduct longitudinal studies across specified areas. Congressional guidance through the budget process—requiring no statutory changes—could clarify priorities and expected frequencies for data collection, as well as provide explicit permission to adjust timelines when needed, and signal when priorities shift.

The recommendations of the NCES leaders we spoke with align well with ASA’s own recommendations to the department. We stand ready to work with the department, the administration, and the education community to revive and improve NCES and its important work.