Why do economists adhere to the practice of alphabetical author ordering? Theory and Experimental Evidence

How to cite

Fan, Simon. 2026. "Why do economists adhere to the practice of alphabetical author ordering? Theory and Experimental Evidence." AEA RCT Registry. AEARCTR-0018670. https://doi.org/10.1257/rct.18670-1.0

Download: BibTeX · RIS — for reference managers without a browser translator. Zotero and most others pick up this page's citation metadata directly.

Simon Fan*

* Primary investigator

Registry ID
AEARCTR-0018670
DOI
https://doi.org/10.1257/rct.18670-1.0
First registered
Last updated
Status
Completed
Trial dates
2025-05-02 – 2025-12-31
Countries
Hong Kong (Hong Kong)
JEL codes
D31; H41; J33
Registry record
http://www.socialscienceregistry.org/trials/18670

Abstract

Alphabetical authorship has long been the prevailing convention in the field of economics. However, despite some papers addressing this topic, there is limited theoretical research dedicated to it. This project aims to conduct a comprehensive theoretical and empirical investigation into the subject. Our main argument is that alphabetical authorship provides a significant advantage to all authors involved in a joint paper. Reversing the alphabetical author order, although potentially beneficial to the first author, may be perceived as unfair to other authors. Consequently, other authors might contribute less to the paper, even when their costs are minimal. This situation could lead to the first author bearing the majority of the workload, which may negatively impact the overall quality of the work. Therefore, individuals may often find it in their best interest not to reverse alphabetical author ordering, even if they possess the initial idea of a project and have sufficient bargaining power to do so. We will test our theoretical predictions through laboratory experiments.

Intervention

We plan to conduct our laboratory experiments using university students in China, a widely adopted approach in related literature. Participants will be recruited through university bulletin board systems and online posters.

Experimental design

Based on what is described in "Randomization Method" below, this project will develop a two-person multi-period strategic interaction model, derive the Nash equilibrium solutions, and investigate the technical conditions under which the following four hypotheses hold.

Hypothesis 1: Compared with A, B is more likely to choose to write alone (i.e., not invite a collaborator in Stage 2).

Hypothesis 2: When B has the idea and invites A to be the second author, A may reject the collaboration even if A’s outside option payoff is lower than the payoff from collaboration.

Hypothesis 3: When B has the idea and invites A to be the second author, A may choose to put in no effort in the co-authorship even if A has higher productivity.

Hypothesis 4: When A’s outside option payoff is lower than A’s payoff from being the second author, B may choose to list the authorship in alphabetical order.

Moreover, we plan to have the following additional treatments.
1. Player chooses authorship ordering & effort is costly
2. Player chooses authorship ordering & effort is costless
3. Authorship ordering is randomly determined & effort is costly
4. Authorship ordering is randomly determined & effort is costless
5. Player chooses authorship ordering & effort is costly & authorship is being observed by third party.
6. Player chooses authorship ordering & effort is costless & authorship is being observed by third party.

Design details

Randomization method
randomization done in office by a computer. Consider a simplified scenario of strategic interaction in a scientific research project involving two individuals (denoted as A and B), where one of them proposes a research idea. The proposer can choose to work alone or invite the other to collaborate. If working alone, the proposer receives 100% of the research output. If collaborating, the distribution of output depends on the authorship order. After the project begins, each individual is endowed with the same amount of time, which can be allocated to work on the project or to other activities (i.e., each has an “outside option”). The final quality of the research output is influenced by the time投入 (input) of both parties. Each individual’s material payoff is positively related not only to the quality of the research output but also to the utility generated by the outside option. The theoretical model consists of a benchmark model and an extended model. 3.1.1 Benchmark Model The basic sequence of the game in the research project between A and B is as follows: Stage 1: Nature determines whether A or B has the project idea and endows both A and B with the same amount of time. Stage 2: The proposer decides whether to write alone or invite the other to co-author. If writing alone, the game ends; the sole author chooses time input and fully owns the output. If inviting co-authorship, the game proceeds to the next stage. Stage 3: The proposer decides the authorship order. Stage 4: The invitee decides whether to accept. If accepting, the two co-author the paper and the game proceeds to the next stage. If rejecting, the invitee allocates all time to other activities, receiving the material payoff from the “outside option” (assumed to be a fixed, relatively low level). Stage 5: Nature assigns one individual a higher productivity and the other a lower productivity. Each individual chooses their own time input to the project. Time not spent on the project can be used for other activities, yielding corresponding material payoffs from the “outside option.”
Randomization unit
Individual
Clusters
400 students
Observations
4000
Arms
4000
Primary outcomes
Predictions: 1. There is a dilemma on effort choice and authorship ordering. A subject who is assigned to determine the ordering may prefer to put herself as first author, however she may afraid the other author to reciprocate by setting a low or even zero effort level. In this case, the author may signal the kindness by choosing to be the second author. (this is a coordination problem) 2. When the authorship ordering is randomly determined, the social welfare (total payoffs) will be higher, and players are more likely to choose high effort. 3. When effort is costless, players will be more likely to choose high effort. 4. When the player A chooses the ordering leading to 40% share for player B, player B will be less likely to choose high effort despite it is socially optimal to do so. 5. When authorship is being observed by third party, players are more likely to choose alphabetical ordering.

Cited by

No citation data available for this registration yet.

Keywords