HERD 2019 Workshop Summary on CapEx and Research Personnel

Attachment 11 - HERD 2020 Workshop Summary.pdf

NCSES - Higher Education R&D and FFRDC R&D Surveys - FY 2020

HERD 2019 Workshop Summary on CapEx and Research Personnel

OMB: 3145-0100

⚠️ Notice: This form may be outdated. More recent filings and information on OMB 3145-0100 can be found here:

2022-06-23 - Revision of a currently approved collection

Document [pdf]

Download: pdf | pdf

Higher Education Research and Development (HERD)
Survey Workshop:
Capital Expenditures and R&D Personnel
September 16–17, 2019

Summary and Next Steps
October 2019

prepared for
National Science Foundation
National Center for Science and Engineering Statistics
by

ICF • 530 GAITHER ROAD • ROCKVILLE, MD 20850

Contents
Background and High-Level Goals ............................................................................................................ 1
Capital Expenditures................................................................................................................................ 2
Capitalized Equipment from Current Funds.......................................................................................... 2
Identifying and Removing Expenditures for Depreciation ..................................................................... 3
Capital Projects.................................................................................................................................... 4
Response to Sample Questions Asking for All CapEx ............................................................................ 5
Equipment-only Awards ...................................................................................................................... 5
Large Purchases Not Capitalized .......................................................................................................... 6
R&D Personnel ........................................................................................................................................ 6
R&D FTEs ............................................................................................................................................. 7
Demographics for R&D Personnel ........................................................................................................ 7
Next Steps ............................................................................................................................................... 8
Capital Expenditures: Collecting all capital expenditures for R&D .......................................................... 8
Capital Expenditures: Removing depreciation ...................................................................................... 9
Capital Expenditures: Reporting Equipment.......................................................................................... 9
Research Personnel: FTEs .................................................................................................................. 10
Research Personnel: Demographics ................................................................................................... 11
APPENDIX A: CapEx on the HERD and Facilities Surveys ......................................................................... 12
APPENDIX B: Workshop Agenda ............................................................................................................ 14
APPENDIX C: Worksheets ...................................................................................................................... 18
APPENDIX D: Capital Expenditure Sample Questions ............................................................................. 23

The National Center for Science and Engineering Statistics (NCSES) within the National Science
Foundation (NSF) hosted a workshop on September 16–17, 2019 at the NSF building in Alexandria,
Virginia. NSF invited representatives from ten universities to an in-person session with NCSES staff to
talk about the reporting of capital expenditures for research and development (R&D) and R&D
personnel data on the Higher Education R&D Survey (HERD).
After some opening remarks from Emilda Rivers, the NCSES director, the university participants and
attendees from NCSES and ICF introduced themselves. The workshop participants represented a diverse
set of universities that complete the annual HERD survey, including four private and six public
universities ranging in size from close to $7 million in annual R&D expenditures to over $800 million. The
participants all contributed to the completion of the HERD survey, some as preparers and some as
reviewers. Three had only been involved in the completion of the survey for one to three years, two had
contributed to the completion of the survey for 4-9 years, but five participants had been involved with
the completion of the survey for 10 or more years. Participants came from various university
administration offices including the offices of research administration, sponsored projects, institutional
research, and controller.
John Jankowski, (NCSES Director of the R&D Statistics Program) and Michael Gibbons (NCSES project
officer for the HERD survey) led the workshop. Michelle Heelan from ICF, the contractor that
administers the survey on behalf of NCSES, facilitated the meeting, and Kathryn Harper (ICF project
director) participated in the discussion.

Background and High-Level Goals
After introductions, Mr. Jankowski provided some background information and high-level goals for the
workshop. He briefly reviewed some of the R&D statistics collected at NCSES, specifically noting the
collection of statistics across economic sectors and how R&D statistics are used by policy-makers. He
then introduced the Frascati Manual. 1 The manual is published by the Organisation for Economic Cooperation and Development (OECD) and is the internationally recognized methodology for collecting
and using R&D statistics. Mr. Jankowski explained that one of the drivers of the workshop was to make
the collection and reporting of R&D statistics in the United States comparable to international statistics,
as guided by the Frascati Manual. He said that there are many topic areas where the United States leads
in providing the guidance for what is collected and how it is collected, but there are still areas without
uniform reporting worldwide. One of the areas where there is the greatest divergence between the
HERD survey and the Frascati Manual guidance is the treatment of “capital use for R&D.” He noted that
the United States is the only country in the world that does not follow the guidelines in the Frascati
Manual for reporting capital use for R&D. He clarified that in the guidelines, the measurement of higher
education R&D includes operating expenses for R&D and capital R&D expenses (e.g., land, buildings,
major pieces of equipment), but specifically excludes depreciation. He noted that depreciation is
currently included in HERD as a component of indirect costs. The workshop will help NCSES determine
whether it can collect additional information that makes higher education statistics more internationally
comparable.
OECD (2015), Frascati Manual 2015: Guidelines for Collecting and Reporting Data on Research and Experimental
Development, The Measurement of Scientific, Technological and Innovation Activities, OECD Publishing, Paris,
https://doi.org/10.1787/9789264239012-en.

Mr. Jankowski also told participants that the purpose of this workshop is just to gather information and
gain a better understanding of the relations between data currently reported on HERD and international
guidelines. He assured participants that the discussions at this workshop would not be the final
determinant that any specific change to survey would be made. This was a message reiterated
throughout the workshop. Participants should not consider any discussions in the workshop to be
evidence that a change to the survey is going to be made. Mr. Gibbons added that there would be
additional interviews conducted with survey respondents before deciding on a course forward.
Mr. Gibbons discussed the types of capital expenditures reported on the HERD survey and the Survey of
Science and Engineering Research Facilities (hereafter referred to in this document as the Facilities
survey). The Facilities survey is another survey of universities and colleges directed by NCSES. The survey
is conducted biennially and collects data on the amount, construction, repair, renovation, and funding of
research facilities. Mr. Gibbons distributed a document summarizing the capital expenditures (CapEx)
collected on both surveys, highlighting the CapEx components not measured on either survey (see
appendix A). The HERD survey collects expenditures on capitalized R&D equipment purchased from
current operating funds in R&D accounts. The Facilities survey collects expenditures or projected
expenditures from capital project accounts for 1) repairs and renovations and 2) new construction of
science and engineering research space. Mr. Gibbons highlighted capital expenditures for R&D not
currently collected on either survey, including land purchased for R&D use, purchased research space,
expenditures for equipment-only awards, and expenditures of intellectual property.
Mr. Gibbons and Mr. Jankowski also prefaced a topic that will be discussed later in the workshop: “Are
there equipment costs that are too large to be included on the HERD survey, considering that
expenditures come from an institution’s current operating funds?” Mr. Gibbons gave some examples of
very large expenditures reported on HERD in recent years, including a research vessel, a supercomputer,
and a giant telescope.
Before proceeding further, Dr. Heelan reviewed the broad goals included in the agenda (see appendix B)
and invited the participants to review the more detailed goals. She then went over the agenda,
explaining that some of the longer sessions would be broken up by paired activities, where partners
would work together to complete a worksheet (see appendix C for all worksheets) and share their
experiences. She explained that the pairs would be asked to share with the group, but there would also
be larger group discussions.
Some topics were discussed multiple time during the course of the workshop. In this summary report,
we attempt to summarize all of these conversations under unified headings, rather than in chronological
order.

Capital Expenditures
Capitalized Equipment from Current Funds
Before addressing questions about how measurements of capital expenditures could be expanded,
participants were asked to share their process for reporting R&D expenditures for capitalized equipment
on the most recent HERD survey, including:
•

At what point, and by whom, was an expenditure, or the project associated with that
expenditure, categorized as R&D?
2

•

•
•

Are internally funded R&D expenditures for capitalized equipment included in the HERD survey?
How is the process for identifying those expenditures different than the process of identifying
externally funded R&D equipment?
How easy or difficult is it to account for capitalized equipment?
How confident are you that all capitalized equipment is accounted for?

Participants worked in pairs to answer these questions and then shared with the larger group. Most
participants described a process where expenditure data were pulled directly from an accounting
system where there is a function code indicating that a particular account is for R&D and another code
for capitalized equipment. One of the smaller institutions, in terms of R&D expenditures, had a more
manual process. Participants felt confident that they were accurately reporting capitalized equipment
purchased as part of a sponsored project because those projects were coded as R&D at the time of
setup. Several participants felt that internally financed equipment purchases might be underreported,
because they were purchased from accounts not designated solely for R&D. They may not have been
coded as R&D because it was a multi-function account or there was an error at the time of setup. This
was complicated for some institutions that had to combine data from different units (e.g., university
foundations) to complete the survey.

Identifying and Removing Expenditures for Depreciation
Institutions were also asked how they would identify and report the amount of indirect costs reported
on the HERD survey that were for depreciation. This is a task that participants had no experience with, in
the context of the HERD survey. Currently, depreciation is included on the survey as a component of
indirect costs; it is not separately reported. Additionally, the HERD survey asks that indirect costs only be
reported for externally funded R&D, not institutionally financed R&D. Again, participants were asked to
work in pairs to share their hypothetical process, consider how difficult it would be to implement, and
express how confident they would feel about the reported value. Participants then shared their
discussions with the larger group.
When the topic of depreciation was first discussed after the first round of paired work, most participants
said that they would start with the depreciation amount for the reference fiscal year, which is typically a
single year-end expense that is charged to one fund code. Some participants did not know how they
would breakdown that one number into specific depreciation amounts for specific R&D assets; this was
usually because they were not personally involved in asset management and depreciation. Other
participants suggested using the depreciation schedule, but they acknowledged that matching up the
plant fund records (where depreciation is tracked) with current-year expenditures would be a
potentially difficult manual process.
Others suggested that because the indirect costs reported on HERD are based on the negotiated F&A
rate, each institution would have to refer to their F&A proposal and determine the portion of the
indirect costs detailed in the proposal that were for depreciation. Related to that, participants noted
that the depreciation included on HERD has no relationship to the actual depreciation expense on R&D
equipment in the current fiscal year. The indirect costs reported on the HERD survey are calculated
based on the institution’s negotiated F&A rate, which may have been finalized three or more years ago.
The topic of depreciation was revisited a few times during the course of the workshop. In early
conversations, participants were not sure how the calculation of the deprecation part of indirect costs
3

would be accomplished, but they thought that they might be able to create a new process. However, as
the conversations continued and participants shared potential hurdles, the path to calculating and
removing depreciation from indirect costs became more complex and burdensome, and the perception
from university participants indicated that numbers would be unreliable.
Throughout the workshop, as the discussion moved to new areas of capitalized expenditures, Mr.
Jankowski returned to the topic of depreciation to better understand the relationship between
deprecation and what is currently reported on the HERD or Facilities surveys, or what might be collected
on the surveys in the future.
•

•

•
•

Mr. Jankowski asked if there may be some over-reporting of capital equipment expenditures
from current funds over time, and participants agreed, but they thought that the amount of
error was very small and the methods of estimating might differ greatly due to the varying
systems and capabilities of individual institutions.
Participants considered the question: “If we had a real annualized amount spent on buildings
and major pieces of equipment, all for R&D, over 15 years, would we get the same total if we
got rid of depreciation and just asked for all R&D equipment and construction/project costs
separately?” The participants pointed out that the F&A proposal is a look back at prior year
depreciation charges. It has no relevance to current depreciation. However, in the aggregate
there would probably be some way to estimate how much bias there is across all the institutions
to be able to make a statistical adjustment.
Participants also considered the question: “Are there any buildings and major pieces of
equipment that are being depreciated, and therefore included in reported indirect costs, that
would NOT be considered R&D as collected in one of the surveys?” A participant responded that
depreciation is based on space usage. If space isn’t used for R&D, it won’t be depreciated.
Mr. Jankowski also asked if offices for research administration would be considered R&D space.
The group agreed that it would not.
Mr. Jankowski asked if depreciation calculations on the F&A proposal included any facilities that
aren’t really for R&D use? Everyone agreed that any facilities included on the F&A proposal
would be for R&D.

Capital Projects
During another paired work session, participants were asked to discuss their process for tracking capital
projects. All participants said that there was a set definition at their institution to determine if a project,
internally or externally funded, was a capital project. In a later discussion, participants said that this is
based on the nature of the final deliverable. When asked how they made sure those projects were not
included on HERD, all participants stated that projects are created in completely different funds (e.g.,
plant funds) that do not intermingle with current fund accounts. Several institutions reported having
capital projects that were classified as R&D (i.e. would be on the Facilities survey, but not on the HERD
survey). Institutions with less R&D did not have R&D capital projects. Lastly, institutions were asked if
they had externally sponsored capital projects. Ms. Harper offered the example of collaborative projects
between two or more universities where the state or federal government might have provided initial
capital as part of a government initiative. Participants noted that most capital projects are financed
through institution funds, but if there was a sponsored capital project, it would go through an approval
process and then be identifiable and excluded from the HERD survey.
4

Response to Sample Questions Asking for All CapEx
During the workshop participants were asked to provide feedback on two sample questions that asked
for all capital expenditures (e.g., land, building, equipment) (see appendix D). Participants first
responded to Sample Question 1, which asked for capital expenditures for R&D in five categories: land,
building, machinery and equipment, capitalized software, and intellectual property products. All
participants said that their institution did not track intellectual property and therefore it could not be
reported.
Participants also thought that land for R&D would be very difficult to report because they rarely buy
land knowing what it is going to be used for. A participant also noted that an institution might purchase
land for R&D but will have to wait a few years for funds to be available to build the research building onsite. She asked when those expenditures should be reported, when you buy the land or when you put it
into service.
A few participants noted that to determine the R&D expenditures for mixed-use buildings the total cost
of the building would have to be prorated by the percentage of space used for R&D. That percentage
would be based on the space survey conducted by each institution in preparation for the F&A proposal.
Mr. Gibbons reminded the participants that building expenditures are already collected on the Facilities
survey, although there are some limitations on the data due to limited reference periods and thresholds
for reporting. He said that NCSES is trying to understand the overlap and missing expenditures between
the two surveys.
A couple workshop participants mentioned that they do not track capitalized software, so they can’t
report expenditures in that category. In a follow-up question, Mr. Jankowski asked if there are
expenditures for R&D software missing from the HERD survey? One participant said, “There could be
some from internal funding that we don’t identify as R&D, but anything that is marked as research we
would pick up.” The participants were asked if there could be a capitalized software purchase that
would not be reported on the HERD survey because it was part of a capitalized project account. Some
participants thought that could occur.
Later in the workshop, participants were asked for feedback on another sample question, Sample
Question 2, which asked for all CapEx by source of funding (federal/nonfederal) and field or R&D.
Participants said that it would be difficult to split building and land expenses across fields. A smaller
institution said that it could be done solely because the participant can sit down with deans to do
allocations, an option that large institutions might not have. Mr. Jankowski noted that the schools
already report expenditures by field, and presumably they could use the same approach to report all
CapEx. Participants said “no,” because currently on the HERD survey most things are on projects
associated with specific faculty members, and that would not be the case with capital projects.

Equipment-only Awards
In pre-work completed by participants prior to the workshop, they were asked if they excluded
equipment-only awards. Some said “yes” and some said “no.” Dr. Heelan asked how the institutions that
excluded those expenditures knew to exclude those funds. All institutions that excluded the funds said
that they knew to do that because the survey review team identified those expenditures and asked for
their removal, explaining that those awards are not for research “activity.”

Dr. Heelan then asked if the exclusion of those awards was appropriate. After a lengthy discussion, all
participants agreed that very large expenditures (whether just for equipment or not) should not be
excluded based on a cap or a percentage of equipment. Most participants acknowledge the spikes that
this might produce in trend data (“lumpiness”). Participants noted that if the same equipment
purchased on an equipment-only award were included in an award with even a small amount of
“activity” included, the expenditures from that award would be permitted on the HERD survey. One
participant suggested adding a question asking for the amount of equipment expenditures that were for
equipment-only awards. She and other participants projected almost no burden involved in that
question, and the data from the question would explain large spikes in equipment expenditures.
Participants noted that large increases should instigate survey staff reviews (as they already do) to
ensure they are being reported appropriately.
A participant suggested that NCSES seems to be doing two things with the same number, “one is using
R&D expenditures as a proxy indicator of research activity, and the other is investment in research.”
“You are trying to accomplish multiple goals. There is no one technical solution that is going to meet
that goal.” It was further suggested that CapEx be collected separately from other R&D expenditures.

Large Purchases Not Capitalized
As part of the effort to learn more about large spikes in expenditures reported on the HERD survey and
ways to address them, participants were asked, “If your institution was awarded the lead on a large
grant to build a center for an off-site research consortium that would be owned by a new nonprofit, how
would the affiliated expenditures be reported on the HERD and/or Facilities surveys?” During the
discussion, several institutions talked about the collaborations they have with other institutions where
the two institutions share research space and split operational costs. There was a lot of discussion about
how to manage ongoing costs, but when one participant pointed out that they had not discussed how to
report the initial large investment, participants either did not have a suggestion or said that they would
just report the expenditures in total as current fund expenditures. It was again suggested that CapEx be
collected separately from other R&D expenditures.
Mr. Jankowski asked how expenditures would be reported if the construction was managed by a
subawardee. Two participants said that it would be reported as passthrough expenditures. When asked
if they would do it another way, no one responded.

R&D Personnel
The second day of the workshop focused on the reporting of R&D personnel. Mr. Jankowski provided
some background on the day’s discussion. He said that the HERD survey collects headcounts of R&D
personnel paid from salaries reported as part of R&D expenditures. The HERD survey asks for the
number of private investigators (PIs) and “other personnel.”
The Frascati Manual has a full chapter on measuring R&D personnel and measuring FTEs, headcounts,
and costs. R&D personnel is comprised of three categories: researchers, technicians, and R&D support
personnel (those integral to the effort but not lab workers). The manual also says that countries should
report the Full Time Equivalents (FTEs). NCSES uses PI headcounts as a proxy for “researchers.” NCSES
does not have any FTE counts for R&D personnel for the higher education sector. NCSES collects FTEs
and headcounts for the business sector for all three categories. The United States is the only country
that does not report FTEs in the higher education sector in the Main Science and Technology Indicators
6

produced by OECD, which includes data from member countries and observer countries. He said that
the workshop will help NCSES better understand its current data, and the potential to derive FTEs for
these researchers, technicians, and R&D support staff.

R&D FTEs
During a paired work session, participants were first asked how the headcounts for PIs and other
personnel paid from the R&D salaries, wages, and fridge benefits were determined when they
completed the FY 2018 HERD survey. Most of the participants mentioned querying the payroll system
for individuals paid from R&D project accounts. Typically, a code identifying the person as a PI is pulled
from another system and matched to payroll. Any individual who was not a PI is counted in “other
personnel.” For some institutions this information can be easily pulled. For other institutions,
particularly smaller ones, the process takes more manual effort.
Two institutions download payroll from sponsored projects. In both of those cases, they do not capture
anyone paid from institutionally financed research. Ms. Harper asked if the institutions that reported
headcounts for internally funded research personnel had internally funded PIs or if “PI” is a designation
limited to sponsored projects. Only one institution had internally financed PIs. At that institution, a
separate account is set up for all projects and the person with the authorization to approve
expenditures on that account is the PI.
Several institutions thought that they were underreporting the headcount of their PIs because the
institutionally financed portion of their salaries for R&D is not separately accounted for and can
therefore not be reported as R&D expenditures.
After describing their current process for reporting headcounts for R&D personnel, participants were
asked how the process would change if they were asked to report FTEs. Participants believe they could
get the necessary raw data. Most mentioned using effort reports and merging that with payroll
information. Most participants also mentioned having to consult with the Office of Human Resources
(HR) on what was possible. The biggest concern for participants was not knowing how NCSES would
want them to calculate an FTE. The unit of time they use to measure an FTE is not typically twelve
months. It might be the length of an academic year (10 months or 9 months), or it might be only be a
few months representing an academic quarter. An example was provided of a full-time faculty member
who also gets a salary during the summer for research performance, which would be calculated as a
1.33 FTE. Additionally, students who received stipends or assistantships to conduct research would not
necessarily have a standard or expected number of hours, so calculating FTEs for students would require
some assumptions.
When asked if they could split “other personnel” (i.e., non-PIs) to pull out students, some participants
said that it would be difficult because there aren’t consistent job titles used in their HR system. They
could not easily identify which job titles belong in each category.

Demographics for R&D Personnel
For the last topic discussed at the workshop, Dr. Heelan asked participants about the feasibility of
collecting demographics for R&D personnel. All institutions said that it would be relatively easy, except
perhaps for graduate students because they don’t report that. Schools report demographic data of
faculty and staff on IPEDS. It is also published in university fact books. They would have to work with the
Institutional Research Office to get those data, but they didn’t think it would be a problem. They would
7

have to provide IR with the list of R&D personnel to look up in their records. If NCSES is interested in
compensation, the data could easily be provided by public universities where that data is public record.
The private universities thought that the provost would probably not approve that.

Next Steps
Below we recommend some possible next steps towards the collection or estimation of HERD data in
the key topic areas discussed at the September 16-17, 2019 workshop. For measures of R&D personnel,
the process of developing survey questions can move forward with approval from NCSES. However,
some details of question will require feedback from NCSES to make sure the needs of data users within
the Center are met. We recommend only a few minor changes to the HERD questionnaire related to the
collection of capital expenditures at this time. However, we do recommend areas for additional
investigation or decision making that could result in more extensive changes to the data collection.

Capital Expenditures: Collecting all capital expenditures for R&D
During the workshop, participants were asked for feedback on two sample questions that asked for all
capital expenditures for R&D (current funds and capital projects) in a specific fiscal year. The first
question asked for capital expenditures by category of expense (land, buildings, machinery and
equipment, software, and intellectual property). The second question asked for expenditures by source
for funds (federal and nonfederal) and major fields of R&D (e.g., engineering, life sciences, and social
sciences). None of the institutions that participated in the workshop tracked intellectual property
purchases and they did not know how they might find that value.
Additionally, the workshop participants had very little confidence in their institution’s ability to identify
and report current-year R&D expenditures for land or buildings. The biggest concern was the
institution’s inability to identify land and building spaces, or portions thereof, as being for R&D (versus
instruction or administration) in the year the expenditure occurred. According to workshop participants,
land is frequently purchased long before it goes into service. Similarly, the expenditures for building
construction can spread across a few years, and the portion of the building that will be used for R&D
may not be known for a few years after the construction begins. Additionally, workshop participants did
not think that they would be able to report these same expenditures by fields of R&D for the same
reasons; they wouldn’t know the fields of R&D conducted on the land or in the building until the
building was in service and a space survey was done.
To collect expenditures from capital projects for R&D use, NCSES will likely have to rely on retrospective
survey questions, which ask for capital project expenditures from 2-4 years in the past, or design a
questionnaire where respondents are comfortable reporting rough estimates for the most recent fiscal
year. After a few years of data collection, past years’ R&D expenditures from capital projects could be
used to derive valid estimates of national expenditures in the most current fiscal year. The Facilities
survey already has a retrospective aspect to it, and given that respondents on that survey already report
construction/renovation costs for R&D by field, we suggest doing additional interviews or workshops
with Facility survey respondents to see what they think might be possible to collect and how to clearly
define it for accurate reporting. Alternatively, questions asking for rough estimates of current year
capital project R&D expenditures, or more precisely measured expenditures from 3-4 years in the past,
could be tested with HERD respondents. The important thing would be to distance, on the

questionnaire, the data collected for capital project expenditures from the highly precise current year
fund expenditures that HERD respondents are committed to provide.

Capital Expenditures: Removing depreciation
Workshop participants did not have clear suggestions about how to exclude depreciation for indirect
costs, at least not within the current structure of the HERD Survey. They agreed that double counting of
research equipment expenditures over a multi-year period was possible, but did not think that there
was a way to appropriately remove depreciation from a report of current-year expenditures. During the
workshop, participants said that the depreciation included in reported indirect costs had no relation to
the equipment cost reported in the current year.
We have two suggestions for how NCSES could proceed. First, to gauge the level of double counting of
equipment expenditures within the population, NCSES could conduct a separate study to 1) learn more
about what types of equipment expenditures are included in F&A proposals (the source of depreciation
expenditures in the HERD survey) and 2) measure the amount of double counting between equipment
included on F&A proposals and equipment reported on the HERD survey for representative sample of
institutions. NCSES could contract with an accounting firm that specializes in assisting universities with
F&A proposals. In addition to providing professional guidance on the types of expenses included in
depreciation estimates, the firm could perform a type of audit with sampled institutions where they
compare the depreciated equipment included it the F&A proposals to equipment expenditures reported
in recent years of the HERD survey. To make sure the results are timely and accurate, NCSES would want
to sample universities or colleges that completed their F&A proposal over the past year.
Alternatively, NCSES could reconsider how indirect costs are reported on the HERD survey. The reason
institutions have a hard time removing depreciation from indirect costs is because the indirect cost is
based on an outdated percentage. If schools were instead asked to estimate current-year actual indirect
cost, not as percentage but as an actual dollar amount, they could then remove the amount for
depreciation without feeling like they were comparing apples to oranges. This would be a significant
change for the institutions and would likely be burdensome.

Capital Expenditures: Reporting Equipment
In addition to measures of depreciation and total capital expenditures, NCSES asked participants about
some capital equipment reporting that might be producing survey error, or, at a minimum, concern
among university data users. Specifically, universities were asked about very large equipment
expenditures that were creating peaks in longitudinal data. The expenditures were not for capital
projects, which are excluded from HERD, but had expenditures that size of a capital project. Although
participants were sympathetic to the “lumpiness” these very large expenditures created in the trend
data, they did not think establishing a cap to limit the size of individual expenditures reported was
justified. As long as an expenditures is for R&D and is from current fund accounts, it should be permitted
to be reported. Participants for larger institutions thought a cap would impact them unfairly.
Participants suggested that NCSES follow up with institutions reporting these large expenditures, as
NCSES has been doing, to verify that expenditures are being properly included in HERD and to get as
many details as possible about the nature of the expenditures.
Participants were also asked about equipment-only awards. NCSES currently asks that respondents not
report equipment-only awards because those awards are not considered R&D accounts, since they do
9

not included R&D activity. However, the guidance has only been provided to institutions in the context
of data review follow-ups or in a webinar in 2018. There is nothing on the HERD survey instructing
institutions to exclude equipment-only awards. Workshop participants did not think those awards
should be excluded. Participants noted that if the same equipment purchased on an equipment-only
award were included in an award with even a small amount of “activity” included, the expenditures
from that award would be permitted on the HERD survey. We recommend for the FY 2020 survey that
the exclusion of equipment-only awards ends; it is hard to justify the continued exclusion of these
awards when very large equipment expenditures continue to be reported.
To address “lumpiness” in the data, participants suggested pulling capitalized equipment expenditures
out of R&D activity expenditures and conducting a separate collection for CapEx. Although a survey or
survey module that collects all CapEx (current fund and capital projects) may be at least a couple years
away (see discussion above), removing capital equipment expenditures from the R&D activity total on
the current HERD survey would be relatively simple. On the HERD questionnaire it would only require
some instruction changes and the removal of capital expenditures from Question 12 of the survey.
Pulling capital expenditures out of R&D activity would likely address some of the problems the HERD
survey has experienced with large spikes in expenditures due to very expensive equipment included in
the R&D project or equipment-only awards. However, this change would create trend breaks in R&D
total expenditures (e.g., Would the total R&D only be for activity [no CapEx] or would it include all
CapEx?) and another series of data tables. NCSES may want to consider this as an option in future years.

Research Personnel: FTEs
Knowing that this is a high-priority for NCSES, and participants said that it was possible to do—although
it was easier for some institutions than others—we would recommend moving forward with designing a
question to add to the HERD survey asking for R&D FTEs by job type. However, in addition to cognitive
interviews, we would recommend a small pilot study where schools are asked to complete the question
and provide feedback on the process for completing the question, what assumptions they made when
calculating FTEs, and what roadblocks they encountered. The cognitive interviews and the pilot can be
coordinated. For example, participants can give initial feedback to first viewing of the question during a
cognitive interview and then get back to us in a month with the completed question and feedback on
the process.
Several workshop participants asked for some very specific directions on how to calculate FTEs and
classify job types. Although these directions can be drafted using guidelines and other surveys (e.g.,
Frascati Manual and BRDIS), the feedback provided during a pilot tests would likely be necessary to
tailor the directions to the unique circumstance of the U.S. higher education sector.
Lastly, NCSES should begin an internal discussion of whether FTEs need to align with R&D salary, and if
they do not, what restrictions should be put on the inclusion of faculty or staff members as R&D
personnel. During the workshop, several participants noted their belief that the R&D personnel were
being undercounted because they needed to be tied to separately accounted-for R&D expenditures
rather than effort or project reporting. Our understanding of the Frascati Manual is that expenditures
and FTEs do not necessarily have to be associated, at least as expenditures are currently being reported
on the HERD survey.

Research Personnel: Demographics
These were the data that the workshop participants were most confident they could provide. We would
recommend moving forward with designing a question to add to the HERD survey. Workshop
participants recommended modeling demographic categories off the questions they already compete
for IPEDS, assuming that they do not conflict with NCSES standards. Although the instructions on this
question will likely need less detail than what will be needed for FTEs, we recommend cognitive and
pilot testing be done simultaneously with the FTE question. We assume that the job categories
requested on both questions will need to be aligned. For that reason, it would be best to test at the
same time.

After NCSES has reviewed recommendations and chosen priorities ICF will propose a long-term project
schedule to guide management of more detailed tasks completed by ICF or other contractors.

APPENDIX A: CapEx on the HERD and Facilities Surveys

Collecting capital expenditures: HERD vs Facilities Survey
HERD

Facilities

Collects expenditures for R&D activities from an
institution’s current operating funds that are
separately accounted for, including . . .

Collects expenditures (and/or projected
expenditures) for (1) repairs & renovations and
(2) new construction of S&E research space,
including . . .

software purchases (noncapitalized and
capitalized),
capitalized equipment (movable equipment
exceeding the institution’s capitalization
threshold [typically $5K]),
equipment purchased from R&D accounts,
depreciation of capital assets in the F&A
calculations.
Data are for a single fiscal year.

any single field with at least $250K in
expended or anticipated completion costs,
fixed equipment within the research space
(included in overall expenditures),
movable equipment that costs $1M or more
(included in overall expenditures).

Costs are collected for projects started during
two-year periods: most recent 2-years for
expended funds, upcoming 2-years for planned
expenditures. Projects may take longer than 2
years.
Field data are prorated.
Costs for the portion of nonresearch space are
excluded.

What’s missing?
-

Research equipment purchased from accounts that are not specifically for R&D
• Equipment-only awards
• Internal university accounts
Land purchased for R&D use
Purchased research space
Other intellectual property products (purchased patents, long-term licenses, or other intangible
assets used in R&D that are in use for more than one year)
Other expenditures?

File Type	application/pdf
Author	Gibbons, Michael
File Modified	2020-12-08
File Created	2020-12-08