Unequal Rewards and the Direction of Scientific Work: Evidence from U.S. Public Universities

arXiv:2611.16747 · econ.GN, physics.soc-ph, q-fin.EC · Submitted 2026-11-12 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Economics. The day in economics: theory, mechanisms and the evidence beneath them.

Ruth: We're looking at the paper: "Unequal Rewards and the Direction of Scientific Work".

Warren: Here's the gist of the paper. Unequal rewards and the direction of scientific work is investigated by testing whether pay dispersion is associated with different kinds of intellectual exploration, finding that inequality favors conventional knowledge production over disruptive innovation The paper tests this hypothesis with administrative payroll records for 105,050 tenure-track faculty members in 4,839 departments at 93 U.S.

Ruth: public universities from 2004 to 2026, linked to publication histories in OpenAlex This evidence shows that inequality is not only a distributional outcome but also an evaluative structure, shaping which forms of knowledge are recognized, rewarded, and reproduced within organizations and communities Pruning redundant reasoning tokens The core argument posits that intellectual tournaments more readily reward the first kind of contribution than the second Pay dispersion may therefore be associated not only with how much scholars produce but also with what kinds of intellectual exploration are encouraged This is tested by estimating a separate pay equation for each department to remove the variation accounted for by its formal ladder, and then measuring the Gini of the remaining pay which captures the department’s unexplained dispersion The paper finds that departments with more unexplained pay dispersion produce more publications and accumulate more citations, as tournament theory expects However, they also produce research that is less disruptive, less novel, and more conventionally connected to prior literature Furthermore, the individual pay premium above a department’s ladder rewards team size and topical breadth rather than disruption or novelty How the training signal is built The study employs a unique dataset collected and assembled with support from a novel AI-agent pipeline The authors introduce a measure of within-organization inequality defined on the organization’s own terms: the dispersion that remains after each unit’s own reward ladder is estimated This residualized Gini is compared against five dimensions of research output: productivity, citations, novelty, conventionality, and disruption The hypothesis H2 is supported across all three compositional measures because unexplained pay dispersion is negatively associated with disruption and novelty and positively associated with conventionality This pattern suggests that highly novel work, being harder to evaluate at the moment it is produced, is less likely to be rewarded by tournaments Where the gains come from The results indicate that scholars in departments with greater unexplained pay dispersion are more productive and more highly cited This association is consistent with tournament theory, contrasting with early survey-based evidence that suggested lower self-reported productivity The study also shows that the pattern of unexplained pay dispersion is a property of departments rather than simply of the disciplinary categories they belong to The authors demonstrate this by comparing departments only within fields and within universities Alternative explanations The paper explores five alternative explanations for the observed patterns, including team composition, selection at appointment, resource endowment, and the mix of sub-specializations inside departments When ranking scholars within each department by mean team size and dropping the top 10 percent of scholars in each department by team size to rule out large collaborations as a factor, the results survive The paper also examines selection at appointment, finding no statistically significant differences in the pre-hire record for disruption, novelty, citations, or productivity The final conclusion suggests that the operative unit is the department and its dispersion is not designed Discussion The broader theoretical claim is that inequality can carry an innovation cost, meaning rewards depend on signals available at the moment of evaluation Existing metrics are better suited to recognizing work that extends the current order than work that revises it The paper concludes that reward systems may therefore amplify what is already legible while discounting disruptive contributions The relevant comparison is not only a scholar deciding how hard to work, but a department deciding how to reward its members The practical implication is that universities that widen internal pay dispersion in pursuit of research excellence may see gains in the outcomes directly observable, yet these departments are less likely to be home to truly disruptive work The paper also makes two methodological contributions: estimating a separate pay equation for each organizational unit and building an AI-agent pipeline linking two decades of public-university payroll records to OpenAlex at the personal level

Title and authors: Ruth: So we're looking at this paper now titled "Unequal Rewards and the Direction of Scientific Work: Evidence from U.S. Public Universities," and what it's really trying to get at is that unequal rewards aren't just about how much work people do, but *what kind* of work gets rewarded.

Warren: Right, it looks like the authors are testing whether pay dispersion is linked to different kinds of intellectual exploration, finding that inequality actually favors conventional knowledge production over disruptive innovation.

Iris: That suggests that the structure of rewards isn't just a reflection of productivity; it shapes which forms of knowledge organizations recognize and reproduce >

Ruth: Exactly. It moves beyond just measuring output and asks what those outputs *are* when the reward structure is skewed >

Warren: And they test this using payroll records from one hundred five thousand fifty tenure-track faculty members across four thousand eight hundred thirty-nine departments at public universities between two thousand four and two thousand twenty-six linked to publication histories in OpenAlex > <ref:2611.16747#pg1>

Iris: So the scale of the data is pretty massive, and they're looking at how that inequality plays out in practice across these different academic settings >

Ruth: It’s a huge dataset, but the core idea is that this inequality acts as an evaluative structure shaping what gets celebrated within those communities >

Warren: And it sets up a really interesting question about whether the pay gap is just a distributional outcome or if it's also setting up the rules for what kind of research gets done >

The paper's summary: Ruth: The paper finds that departments with more unexplained pay dispersion produce more publications and accumulate more citations, which aligns with tournament theory expectations >

Warren: But here’s the twist, they also found that this kind of department produces research that is less disruptive and less novel, and it tends to be more connected to prior literature >

Iris: So the pattern isn't just about how much research comes out; it's about the *type* of research that gets rewarded when there’s unexplained pay dispersion present >

Ruth: That’s right. They also found that this premium above a department's formal ladder rewards things like team size and topical breadth, rather than novelty or disruption >

Warren: So what they're saying is that the structure of the reward system itself is guiding the direction of scientific work, favoring established paths over risky new ideas >

Iris: And they test this by looking at five dimensions of research output: productivity, citations, novelty, conventionality, and disruption >

Ruth: The whole point is that unexplained pay dispersion is negatively associated with both disruption and novelty but positively associated with conventionality across those measures >

The paper's improvements: Warren: Now for the methodology, they introduce a unique way of measuring inequality called the residualized Gini, which captures the gap between what a person earns and what their rank, contract basis, and seniority predict they would earn in that department in that year >

Ruth: That means they estimate a separate pay equation for each department to take out the variation accounted for by its formal ladder before calculating this residual dispersion >

Iris: And then they compare this residual Gini against those five dimensions of output—productivity, citations, novelty, conventionality, and disruption—to see how the unexplained gap relates to that research profile >

Warren: The hypothesis they test is that unexplained pay dispersion is negatively associated with disruption and novelty but positively associated with conventionality across all three measures >

Ruth: They also found that this pattern holds even after removing big team scholars from every department, meaning the raw pay inequality and the inequality the ladder itself creates carry the same signs >

Iris: And they looked at other explanations for these patterns, like team composition or selection at appointment, and they found no statistically significant differences in those factors >

Conclusion: Ruth: So to wrap up this work on "Unequal Rewards and the Direction of Scientific Work: Evidence from U.S. Public Universities," the main implication is that inequality can carry an innovation cost >

Warren: Meaning rewards depend on signals available at the moment of evaluation, and existing metrics are better suited for recognizing work that extends the current order rather than work that revises it >

Iris: The practical implication is that reward systems might amplify what's already legible while discounting contributions that are truly disruptive >

Ruth: So maybe the comparison isn't just about a scholar deciding how hard to work, but about a department deciding how to reward its members and what kind of scholarship they value >

Warren: It suggests universities that widen internal pay dispersion in pursuit of research excellence might see gains in observable outcomes, but those departments are less likely to be home to truly disruptive work >

Iris: It's an interesting finding about how the structure of rewards can become a constraint on the kind of science being done >

Likun Cao

Purdue University

econ.GN, physics.soc-ph, q-fin.EC

Submitted: 2026-11-12

Updated: 2026-11-12

Comments: 42pages, 4 figures, 3 tables

License: http://creativecommons.org/licenses/by-nc-nd/4.0/

Importance score: 89/100

The gist: Unequal rewards and the direction of scientific work is investigated by testing whether pay dispersion is associated with different kinds of intellectual exploration…

Key concepts

Unexplained Pay Dispersion
This measures the gap between what a scholar actually earns and what their rank, contract basis, and seniority would predict they would earn in that specific department. It captures the part of pay variation not explained by formal departmental structures.
Tournament Theory
This theory suggests that intellectual tournaments—environments where scholars compete for recognition—tend to reward the first kind of contribution (conventional knowledge) rather than disruptive innovation, as these are easier to evaluate and reproduce.
Knowledge Output Dimensions
These are five ways research can be measured: productivity (how much is produced), citations (how often it's used), novelty (how new the idea is), conventionality (how connected it is to existing literature), and disruption (how much it changes the field).
Evaluative Structure
This refers to how organizations decide what kind of work is valued. The paper argues that pay inequality isn't just a result of production; it actively shapes which forms of knowledge are recognized, rewarded, and reproduced within academic communities.

Terminology

Summary

Unequal rewards and the direction of scientific work is investigated by testing whether pay dispersion is associated with different kinds of intellectual exploration, finding that inequality favors conventional knowledge production over disruptive innovation The paper tests this hypothesis with administrative payroll records for 105,050 tenure-track faculty members in 4,839 departments at 93 U. p. 0

Where the paper says it

Pay dispersion may therefore be associated not only with how much scholars produce but also with what kinds of intellectual exploration are encouraged. We test this argument with administrative payroll records for 105,050 tenure-line faculty in 4,839 departments at 93 U.

S. public universities from 2004 to 2026, linked to publication histories in OpenAlex This evidence shows that inequality is not only a distributional outcome but also an evaluative structure, shaping which forms of knowledge are recognized, rewarded, and reproduced within organizations and communities

Pruning redundant reasoning tokens

The core argument posits that intellectual tournaments more readily reward the first kind of contribution than the second Pay dispersion may therefore be associated not only with how much scholars produce but also with what kinds of intellectual exploration are encouraged This is tested by estimating a separate pay equation for each department to remove the variation accounted for by its formal ladder, and then measuring the Gini of the remaining pay which captures the department’s unexplained dispersion The paper finds that departments with more unexplained pay dispersion produce more publications and accumulate more citations, as tournament theory expects However, they also produce research that is less disruptive, less novel, and more conventionally connected to prior literature Furthermore, the individual pay premium above a department’s ladder rewards team size and topical breadth rather than disruption or novelty

How the training signal is built

The study employs a unique dataset collected and assembled with support from a novel AI-agent pipeline The authors introduce a measure of within-organization inequality defined on the organization’s own terms: the dispersion that remains after each unit’s own reward ladder is estimated This residualized Gini is compared against five dimensions of research output: productivity, citations, novelty, conventionality, and disruption The hypothesis H2 is supported across all three compositional measures because unexplained pay dispersion is negatively associated with disruption and novelty and positively associated with conventionality This pattern suggests that highly novel work, being harder to evaluate at the moment it is produced, is less likely to be rewarded by tournaments

Where the gains come from

The results indicate that scholars in departments with greater unexplained pay dispersion are more productive and more highly cited This association is consistent with tournament theory, contrasting with early survey-based evidence that suggested lower self-reported productivity The study also shows that the pattern of unexplained pay dispersion is a property of departments rather than simply of the disciplinary categories they belong to The authors demonstrate this by comparing departments only within fields and within universities

Alternative explanations

The paper explores five alternative explanations for the observed patterns, including team composition, selection at appointment, resource endowment, and the mix of sub-specializations inside departments When ranking scholars within each department by mean team size and dropping the top 10 percent of scholars in each department by team size to rule out large collaborations as a factor, the results survive The paper also examines selection at appointment, finding no statistically significant differences in the pre-hire record for disruption, novelty, citations, or productivity The final conclusion suggests that the operative unit is the department and its dispersion is not designed

Discussion

The broader theoretical claim is that inequality can carry an innovation cost, meaning rewards depend on signals available at the moment of evaluation Existing metrics are better suited to recognizing work that extends the current order than work that revises it The paper concludes that reward systems may therefore amplify what is already legible while discounting disruptive contributions The relevant comparison is not only a scholar deciding how hard to work, but a department deciding how to reward its members The practical implication is that universities that widen internal pay dispersion in pursuit of research excellence may see gains in the outcomes directly observable, yet these departments are less likely to be home to truly disruptive work The paper also makes two methodological contributions: estimating a separate pay equation for each organizational unit and building an AI-agent pipeline linking two decades of public-university payroll records to OpenAlex at the personal level

Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.

Improvements for AI systems

  1. Improved AI can perform residualized pay dispersion estimation by fitting a separate pay equation for each organizational unit to remove variation accounted for by its formal ladder, as described in Section 3.2 of the paper.

  2. Improved AI can measure the Gini of this residual dispersion, which captures the gap between what a person earns and what their rank, contract basis, and seniority predict they would earn in that department in that year, as stated on Page 10.

  3. Improved AI can calculate five knowledge output dimensions for each scholar: disruption (using the three-year Funk–Owen–Smith CD index), novelty (using Uzzi et al. 2013 measures), conventionality, and citations (C3), as detailed in Section 3.3 on Page 14.

  4. Improved AI can estimate standardized coefficients from OLS models that relate unexplained pay dispersion to these five knowledge dimensions, such as the association between unexplained pay dispersion and the dimensions of output that are observable at the time of evaluation (Page 18).

  5. Improved AI can perform a stacked difference-in-differences test by event cohort for retirements and arrivals of top-paid senior members, to observe how these events move the department’s unexplained pay dispersion in opposite directions (Page 27).

  6. Improved AI can analyze the profile of scholars based on their research style, finding that departments reward the scholar who works across many topics and does not specialize narrowly, as summarized in Section 4.4 on Page 23.

Abstract

Research on pay dispersion has long asked whether unequal rewards increase or depress performance. Tournament theory expects stronger effort, while social comparison accounts expect withdrawal. Both perspectives, however, treat performance as a single scale on which workers can be ranked. Scientific work often resists such ordering. Some contributions extend recognized lines of inquiry and can be evaluated against prevailing standards; others revise the frames through which value is judged and become legible only later. We argue that intellectual tournaments more readily reward the first kind of contribution than the second. Pay dispersion may therefore be associated not only with how much scholars produce but also with what kinds of intellectual exploration are encouraged. We test this argument with administrative payroll records for 105,050 tenure-line faculty in 4,839 departments at 93 U.S. public universities from 2004 to 2026, linked to publication histories in OpenAlex. For each department, we estimate a separate pay equation using rank, contract basis, and seniority, then measure the Gini of the salary variation this local ladder cannot explain. Scholars in departments with greater unexplained pay dispersion publish more papers and receive more citations, but their work is less disruptive, less novel, and more conventional. Additional analyses show that compensation premiums above the ladder go disproportionately to scholars with broad collaborations and topical breadth. Although the design cannot fully separate treatment from selection, the evidence shows that inequality is not only a distributional outcome but also an evaluative structure, shaping which forms of knowledge are recognized, rewarded, and reproduced within organizations and communities.

Sources

Related papers