Dynamic investment and coordination in teams: Theory and experimental evidence

arXiv:2611.13230 · econ.GN, cs.GT, q-fin.EC · Submitted 2026-11-09 · 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: I'm Ruth.

Warren: And I'm Warren, here with Iris.

Ruth: We're looking at the paper: "Dynamic investment and coordination in teams".

Warren: Here's the gist of the paper. Teamwork necessitates investments in team-specific skills or capital that accumulate within a team, leading to increased returns on coordinated production.

Ruth: This paper presents the first systematic analysis of dynamic investment in such team skill, establishing a novel theoretical framework where investment carries risk because its benefits are contingent upon successful team coordination.

Warren: The game is inherently dynamic, as each investment generates a persistent increase in team skill that endures throughout the duration of the team relationship.

Ruth: The theoretical framework is structured into two main parts: Part 1 involves players simultaneously deciding whether to invest in team skill, which raises the collective skill level by a fixed amount for all subsequent rounds.

Paper summary: Ruth: So, wrapping up this work on "Dynamic investment and coordination in teams: Theory and experimental evidence," they’re showing that while teamwork is crucial for modern production, people often underinvest in team skill because they don't fully account for how much helping each other actually boosts the outcome >

Warren: The authors use a dynamic game model to show that this investment is risky because you can't guarantee coordination, but the results show that incentives do matter; stronger incentives lead to more investment and better outcomes overall when players have the right traits >

Iris: They pinpoint theory of mind and the Coordination Attraction Score as key predictors for both individual skill investment and team efficiency, which helps explain why some teams perform much better than others based on their makeup >

Ruth: It’s about realizing that the benefit to every team member depends on successful coordination, which means principals need to focus on designing those incentives correctly rather than just assuming everyone will invest optimally >

Warren: The paper’s implication is that if you want real gains from teamwork, you have to think about the structure of the incentive design and how you select your team members based on those psychological traits >

Iris: They also mentioned that while they found these patterns, their method doesn't cover everything; for instance, they didn't explore other settings like prisoner’s dilemma games or public goods games yet >

Ruth: Yeah, so the limitation is that this study focuses specifically on this dynamic investment game and its relationship with coordination >

Warren: It stops short of exploring how these investment dynamics play out in other complex economic scenarios where coordination might be even more complicated >

Conclusion: Ruth: So, we're wrapping up our look at "Dynamic investment and coordination in teams: Theory and experimental evidence." Basically, the paper argues that teams don't just need to work together; they have to actively invest skill into that teamwork, but this investment is risky because it totally depends on everyone coordinating well.

Warren: Right. The core idea is that you can't just put more effort in blindly; you have to coordinate for the returns to even happen, and that coordination risk means you might end up doing something individual instead.

Iris: I looked at the experimental setup, and what’s interesting is how they used theory of mind and a specific score, the Coordination Attraction Score, to predict both how much people invested and how efficient their team actually was. It shows those personality traits matter.

Ruth: Exactly. The paper found that underinvestment in team skill is common even when teams are playing optimally together, just because people don't see the benefit of helping each other out—those positive externalities they get from coordinating.

Warren: And the results show that if you have stronger incentives, which we call b, people invest more in team skill, but that investment drops off as the rounds go on within the same match. It’s a trade-off between getting a good return now and trying to keep building up skill later.

Iris: The authors also found that higher theory of mind ability and that Coordination Attraction Score predict which matches are assigned to certain clusters, meaning you can actually use those traits to sort your team members for better performance. That’s a practical application there.

Ruth: So, the big picture is that we need to design incentives smartly and choose people with the right coordination skills if we want real productivity gains from teamwork in practice. It's not just about having a good idea; it's about how you structure the environment around that idea.

Warren: It really comes down to realizing that without those external pushes or internal skill building, the theoretical benefits of cooperation just stay locked up inside the group.

Iris: And while this study is strong on these experimental findings, they’ve acknowledged that they didn't explore how this whole dynamic investment thing would work in other game structures like Prisoner’s Dilemma or public goods games yet. That leaves some room for future testing.

David Gill, Victoria Prowse, J. Lucas Reddinger

Purdue University · Princeton University

econ.GN, cs.GT, q-fin.EC

Submitted: 2026-11-09

Updated: 2026-11-09

Comments: 90 pages, including appendices. JEL: C73, C92, D91, J24

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 100/100

The gist: Teamwork necessitates investments in team-specific skills or capital that accumulate within a team, leading to increased returns on coordinated production.

Key concepts

Dynamic Investment
Players repeatedly decide whether to invest in team skills, which permanently raises the collective skill level for all future rounds. This process is dynamic because the benefit of an investment persists throughout the entire duration of the team relationship.
Coordination Dependence Risk
The paper argues that investing in team skill carries risk because its positive outcomes are entirely dependent on successful coordination among teammates. If coordination fails, the investment yields little or no benefit, making it inherently risky.
Positive Externalities
When teammates coordinate effectively, their individual efforts produce a greater collective benefit than if they worked alone. This concept explains why players might be motivated to invest in team skill to reap these shared gains.
Theory of Mind (ToM) and CAS
These are individual traits measured in the study that predict player behavior. Theory of Mind relates to the ability to understand others' mental states, while Coordination Attraction Score measures a propensity for coordination. These traits influence both investment decisions and overall team efficiency.

Terminology

Summary

Teamwork necessitates investments in team-specific skills or capital that accumulate within a team, leading to increased returns on coordinated production. p. 0 This paper presents the first systematic analysis of dynamic investment in such team skill, establishing a novel theoretical framework where investment carries risk because its benefits are contingent upon successful team coordination. p. 0 The game is inherently dynamic, as each investment generates a persistent increase in team skill that endures throughout the duration of the team relationship. p. 2

The theoretical framework is structured into two main parts: Part 1 involves players simultaneously deciding whether to invest in team skill, which raises the collective skill level by a fixed amount for all subsequent rounds. p. 2 Part 2 models player choices between undertaking the team task or an individual task, structured as a Stag Hunt game whose characteristics depend on the current level of team skill. p. 7 Key properties identified include: investment risk due to coordination dependence, positive externalities for teammates when coordinating, dynamism from persistent skill growth, and the eventual decline of dynamic benefits as the supergame progresses. p. 2

Empirical investigation involved laboratory experiments with over 600 subjects playing this dynamic investment game multiple times. p. 3 Findings indicated substantial underinvestment in team skill relative to the socially efficient benchmark. p. 3 Investment decisions are highly responsive to incentives; stronger incentives motivate subjects to invest more and achieve greater social efficiency more frequently. p. 3 Pre-registered traits, specifically theory of mind and a propensity to coordinate (measured by the Coordination Attraction Score, CAS), serve as predictors for both investment levels and team-level efficiency. p. 4

Analysis of the dynamic investment game in the laboratory revealed that underinvestment compared to the socially efficient benchmark is consistent with team play subgame-perfect Nash equilibrium (team play SPE). p. 3 In this equilibrium, players coordinate on the efficient team task but underinvest in skill relative to the socially efficient benchmark because they fail to account for positive externalities. p. 3 Investment is inherently risky due to its dependence on coordination, and subjects sometimes opt for the safer individual task instead. p. 3 Investment strongly responds to incentives, and higher theory of mind ability and CAS predict investment in team skill when the team play SPE predicts such investment. p. 0

Machine learning clustering of matches demonstrated that heterogeneity is primarily organized by the intensity and timing of investment: most matches achieve coordination on the team task but vary in the extent and timing of their skill investments, with a smaller subset exhibiting coordination breakdown. p. 4 Average theory of mind ability and CAS predict which cluster a match is assigned to. p. 5

The findings suggest that economic returns from teamwork are critically dependent on incentive design and team composition, implying that principals can exert influence through incentive structuring and member selection based on traits. p. 4 Investment in team skill increases with the level of incentives (b) across matches but diminishes as rounds progress within a single match. p. 3 Subjects possessing higher theory of mind ability and CAS are more inclined to invest in team skill when the team play SPE anticipates such investment. p. 4 The data support hypotheses concerning the effects of b on investment and round-to-round changes in investment, as well as the predictive power of individual traits on both task choices and investment decisions. p. 7 Furthermore, average theory of mind ability and CAS predict match-level efficiency. p. 4

The paper concludes that despite the importance of teamwork, underinvestment in team skill suggests that realizing full benefits is challenging due to uninternalized positive externalities. p. 2 This implies that principals should leverage incentive design and trait-based team member selection to unlock productivity gains. p. 4 Future research plans include exploring strategic incentive design, endogenous sorting mechanisms, and the impact of team size. p. 4 The theoretical framework is also extended to other settings such as Prisoner’s Dilemma or public goods games. p. 5

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. Automated Team Skill Investment Prediction: The AI system can predict whether a subject will invest in team skill based on their predicted coordination level and cognitive ability, as shown by Hypothesis 3 results from Table 6. This allows for proactive intervention or personalized incentive design tailored to individual traits.

  2. Dynamic Incentive Optimization: The system can suggest optimal incentive strengths (varying how much each investment increases team skill) that maximize the probability of achieving full social efficiency in a match, as supported by results in Table 3 and the discussion on incentives in Section V.B.

  3. Team Formation and Composition Guidance: The AI can recommend selecting team members with higher theory of mind and Coordination Attraction Score to maximize the probability of a successful team outcome, based on the findings that these traits predict investment and task choice (Table 7).

  4. Match-Level Strategic Behavior Clustering: By using machine learning to cluster matches, the system can categorize strategic interactions into distinct behavioral types (Clusters 1 through 7), allowing for targeted strategies based on which cluster a match is likely to fall into.

  5. Round-Specific Investment Guidance: The system can determine the optimal timing for investment within a team interaction by predicting when investment will stop, specifically identifying the transitions between rounds where the decline in investment rate is largest (Hypothesis 2a).

Related papers