8  Evaluating the Budget Proposal Realistically

8.1 Statistical Cross-disciplinary Collaboration & Consulting Lab

Let us consider the proposed scenario of retaining SC3L for providing statistical consultation to IANR faculty, staff, post-docs and students. Currently, the SC3L is headed by a director who is a tenured associate professor of Statistics with more than 9 years of experience in consulting. There are 5 Statistics graduate research assistants who directly work with the clients. These GRAs are senior Statistics PhD students who are required to take Stat 821, 822, 882, 883, 825 before they are eligible to receive the GRA; only the best students across these courses are selected to work at the SC3L. Additionally, students enrolled in Stat 825 and Stat 930 play important supporting role in resolving the research problems that are brought to the SC3L. Clearly, the SC3L is a team effort that requires that requires each member to have completed advanced Master or PhD level Statistics courses.

Eliminating the BS/MS/PhD programs in Statistics will also eliminate most of that team. So, the SC3L cannot be retained in current form under the proposed elimination. Note that the MS and Ph.D. Statistics programs were formed by the desire to increase consulting and collaboration capacity within the department while providing students with experiential learning and training in consulting methods. That is, our programs exist in part because of a desire to provide better service to the university.

What other forms of SC3L could be envisioned that operate with similar efficiency?

• Hiring PoPs?
  • Between 2020-2025, the SC3L has served \(\approx 131\) clients per year. This implies that the combined powers of a director and five graduate students can allocate about 3 days to each client. It is not feasible for a single PoP to take on the workload of the entire SC3L team. The number of meetings combined with the amount of non-meeting work (computer programming, checking models, writing reports) would make it impossible for one person to complete this work efficiently.
  • Consequences: The research conducted by IANR faculty, staff, post-docs, and students, as well as other units, will suffer. This will become an impediment to achieving the “Extraordinary Research” ideal of Odyssey to the Extraordinary.
• One PoP and a team of graduate students?
  • The current plan eliminates the statistics graduate and undergraduate degree programs along with all of their core classes. Thus, there will be no graduate students in Statistics to staff the desk, and no courses to train these students (or students from other disciplines). Without training, this version of the SC3L would be ineffective.
  • Could graduate students be recruited from other departments, such as Agronomy? Perhaps, but these students have much less quantitative background, and with the elimination of the statistics graduate program, there is no clear way to provide these students with the training they would need in order to be functional. The quality of consulting available on campus under this plan would be extremely low, and the PoP would be less efficient because of the need to train outside students while handling an extreme load for consulting.
• Two POPs?
  • This configuration has a better chance of achieving the same efficiency as we currently have, if people willing to take the positions can be located. To solve the problems handled by SC3L, we need highly skilled statisticians with enough experience in handling different types of complex data. Such experience can come either through PhD training in Statistics or an extensive stint in industry (or both). A recent Ph.D. graduate is likely to be the most cost effective option. Even in the unlikely event that such a person could be recruited, how much would they cost? The American Statistical Association periodically publishes a salary survey among statisticians. From the last round of survey (circa 2021) it appears that an entry level statistics instructor (analog of PoP) costs approximately 3/4 of the salary of an entry-level assistant professor in statistics. However, someone with this profile is more likely to accept an industry job because the same consulting skills useful in the SC3L are much better compensated in industry. It is likely that it would cost more than 2 FTE from the current department to find someone willing to handle the workload of the SC3L. The net 12 FTE savings is not realistic.

The SC3L will require at least 2 FTE to maintain, even without the associated statistics department programs, leaving a cost savings of 11 FTE relative to the current 13 FTE. It is possible that these PoPs could be moved to soft money positions, like those in Biostatistics programs. However, our department went through a grueling hiring cycle to try to locate a tenured hard-money director of the SC3L and ended up hiring internally instead because we could not attract someone qualified. Based on that experience, we can confidently predict that IANR will not find a trained Ph.D. statistician who is willing to take a soft-money PoP role at UNL. It seems unlikely that IANR will be able to find someone willing to take a PoP role to run the SC3L, given that we could not find someone willing to take a tenured position. If the administration proceeds with this proposed cut in its current form, it seems likely that they will fail to recruit the consulting expertise necessary to even have a statistics presence on campus, leaving researchers in a terrible position. Many grants took for granted the existence of the SC3L and free consulting services, or the willingness of members of the Statistics department to collaborate without being listed as a co-PI on the grant. What will happen to these projects without statisticians on campus?

In truth, it will likely be hard to recruit any statistical expertise to UNL after the publicity surrounding this budget proposal. Statistics is an incredibly in-demand field, so who would risk moving to Nebraska if the university does not understand or appreciate the role of a statistics department within a public land grant institution?

8.2 General Education Courses

Next, who teaches Stat 218, 380, 801, 802, and 870?

Currently, the Statistics department offers approximately 17 sections of Stat 218 (8 in fall, 7 in spring, and 2 in summer), in addition to 6 sections of Stat 380, and two sections of 801 (with 2 sections of lab each) and 802. Stat 870 is offered less frequently, so we will exclude it from this analysis in order to produce conservative estimates. The courses identified to be kept require 25 sections per year across 4 preps; we estimate that this would require at least 3 professors of practice to teach (assuming a 4-4 load) which are not accounted for by the current budget reduction plan. These FTEs would need to be subtracted from the savings listed, yielding only 9 FTE savings for cutting four programs (BS in Statistics, BS in Data Science from CASNR, MS in Statistics, and Ph.D. in Statistics).

We use professors of practice for this comparison rather than adjuncts both because it seems unlikely that 25 courses could be assigned to adjuncts with statistical training, given that most people with graduate degrees in Statistics can make more freelancing as data scientists than they would be compensated for teaching courses. Certainly, it seems likely that the FTEs which are being eliminated will not be available to teach courses at adjunct rates, as many other Big Ten and AAU institutions are hiring, and some institutions have multiple positions¹.

The calculations for how many PoPs would be required to teach current Statistics courses that drive revenue generation for the department does not include essential courses identified by other colleges, such as Stat 462 and Stat 463, which are required for the Actuarial Science degrees in Business and CAS. Two additional Stat 300/400 level courses are required beyond Stat 380 for the Mathematics, Statistics, and Data Science focus area within the Math department. The Digital Agriculture minor also requires Stat 151 and 251, computing courses developed for the Statistics undergraduate major. The Agricultural Economics Ph.D. requires Stat 882, and the Finance Ph.D. requires 9 hours of graduate Statistics coursework (it is quite possible that Stat 882 and 883 would be preferable to 801 and 802 for Finance majors).

This analysis does not consider the fate of the Data Science programs in CAS and COE, which would lose between one and three courses in the data science core as well as the statistics focus area that is primarily made up of courses designed for the Statistics and Data Analytics major and the Statistics minor, as selecting a minimal subset of these courses will still damage the flexibility within the Data Science program and may lead to students selecting other majors rather than Data Science.

In order to support these additional courses, an additional professor of practice would likely be required, reducing the FTE savings from eliminating the department to 9, or 7 if the SC3L is kept as planned.

Calculations	FTE
Current Department FTE	13
FTE required for 218, 380, 801, 802	-3
FTE required for 151, 251, 462, 463, 882	-1
FTE required for SC3L	-2
Total	7

It is important to note that this is not an apples-to-apples comparison. 6 PoPs (2 for SC3L, 4 for teaching) cannot possibly maintain the quality of the educational and research contributions the Statistics department maintains – there is simply not enough time in the day to meet the demands of PoP positions and stay current on new developments in Statistics. The quality of statistical education and consulting under this distributed/PoP/adjunct model will inevitably degrade. Without a centralized unit facilitating course coordination, the variability in course quality will increase, eventually leading to the types of concerns that were raised during the 2005 Statistics APR, where the quality of 801/802 were of concern in part due to the lack of coordination during the upheaval of trying to combine the Biometry and Statistics core within the Math Department.

8.3 Grant Funding Losses

Elimination of the department of Statistics will lead to the loss of already-promised federal funding from NSF and NIH, among others. In total, there is more than $1M of federal grant money that Statistics faculty have as PIs, as of October 2025.
Termination will lead to permanent loss of these research restricted dollars (a vital AAU metric).

It is not possible to calculate the loss of research dollars due to the damage to UNL’s reputation, but several letters in support of the department raise this issue (and the authors of those letters are individuals who have served at high levels within federal funding agencies). Statistical collaborators are a resource just like the Holland Computing Center or the greenhouses and fields available for ag research or the labs and equipment available for other scientific disciplines. UNL has very publicly announced that they do not prioritize this resource, and funding agencies will undoubtedly be aware of this issue when evaluating grants from UNL PIs. In an era where federal agencies are looking for any excuse to deny funds to universities, this is a risky position to take as an institution.

8.4 Tuition Generation

The statistics department had a total realizable base tuition of $3.76 million dollars compared to its state aided budget of $2.54 million. We understand that total realizable tuition is an over-estimate of the amount of tuition collected due to remissions and scholarships. For the sake of argument, suppose that only 70% of possible tuition is collected by the university. The statistics department as-is then generates $2,630,820 in revenue, which is larger than our state aided budget.

8.5 SDAN Undergraduate Major

We expect SCH to increase due to our undergraduate major (Statistics and Data Analytics, or SDAN), particularly after AY 2025-26, because we will graduate our first class of students and be able to advertise based on their successes. We would also expect SCH generation to increase were the SDAN program to be moved into a different college – while anecdotes are not data, we have received considerable feedback that being 1) located on East campus, and 2) having CASNR general education requirements lowers enrollment in our program. Students could also double-major more easily if we were in CAS, COE, or COB. Thus, we would expect that if our department was realigned to a different college, located on City campus, or both, we would be able to generate more SCH and be considerably more profitable as a result.

The creation of the SDAN program represented an investment in Statistics, and that investment has not yet reached its maturity date. It is critically important that UNL stay the course until the first two or three cohorts graduates in SDAN before eliminating Statistics, or it will never realize the return on the investment it made in 2021 when the program was first approved. While the “sunk costs” fallacy is absolutely a concern, UNL has not actually invested additional funding in SDAN beyond a part-time undergraduate advisor and some online advertising in the first two years of the program. The primary investment has been the excess teaching load required to create 16 new courses (42 hours), which necessarily will decrease research productivity in the short term. However, by the end of Spring 2026, all of this work will have been completed, and UNL can sit back and reap the benefits of the department’s labors.

8.6 Conclusion

Eliminating the Statistics department while continuing to teach key courses (both those identified by IANR and those identified by CAS/COB) and maintain even basic SC3L functionality will only save approximately 7 FTE. At least 2 FTE are required for even basic SC3L functions (and this still represents a massive decrease in consulting resources within IANR), and ~4 FTE PoPs would be required to teach the courses.

When considered against both the tuition revenue generated by all stats courses and the losses in grant funding, awards, and prestige, this is not at all a good proposition, particularly considering that collecting even 70% of realizable base tuition is sufficient to make the department profitable.

If UNL then factors in the losses in grant funding due to uncompetitive proposals and lack of statistics collaborators, losses in SCH in Agronomy and other departments due to inability to maintain enrollment in plant breeding programs as a result of the statistics department’s elimination, and losses in enrollment to data science programs in CAS and COE due to the loss of statistics courses and options within those programs, this proposal does not make sense from a budget perspective.

As the undergraduate program matures, we will be able to as a department overhaul the structure of our MS courses, increase the number of majors in our SDAN program through outreach efforts, and fund PoPs for the department through differential tuition to reduce the teaching load on faculty due to the SDAN program. The soon-to-be-proposed online MS in Data Science also represents an opportunity to increase profitability.

---

1. This leads to the ironic possibility that members of the Statistics department may make it into the AAU long before Nebraska does, particularly if this proposal goes through.