Streaming problems as (multi-issue) claims problems

Gustavo Bergantiños · 2024

Read Streaming problems as (multi-issue) claims problems by Gustavo Bergantiños (2024) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2412-18628.txt.

Introduction

Music streaming services take in subscription money and must split it among artists, and Gustavo Bergantiños and Juan Moreno-Ternero show that this modern problem is an old one in disguise. The field of claims problems, which studies how to divide a scarce pot among claimants, already owns the tools. The authors prove that the main streaming methods, the pro-rata rule, the user-centric rule, and families that blend them, each match a known claims rule.

The streaming problem can also span several issues at once, such as different countries or subscription tiers, and the claims framework covers that case too. Their results connect a well-tested body of fairness theory to a young industry debate. Streaming platforms and regulators can now borrow decades of results about which splits reward which behaviour.

Catalog record

CREATORGustavo Bergantiños
YEAR2024
COLLECTIONPAPERS
FORMATTXT
FILE SIZE42 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2412-18628

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Streaming problems as (multi-issue) claims problems Thanks: Financial support from grants PID2020-113440GBI00 and PID2023-146364NB-I00, funded by MCIN/AEI/ 10.13039/501100011033 and MICIU/AEI/10.13039/501100011033/ respectively, and by FEDER, and UE is gratefully acknowledged. Thanks: Department of Economics, Universidad Pablo de Olavide, 41013 Sevilla, España; [email protected] Gustavo Bergantiños Abstract We study the problem of allocating the revenues raised via paid subscriptions to music streaming platforms among participating artists. We show that the main methods to solve streaming problems (pro-rata, user-centric and families generalizing them) can be seen as specific (well-known) rules to solve (multi-issue) claims problems. Our results permit to provide strong links between the well-established literature on claims problems and the emerging literature on streaming problems. JEL numbers : D63, C71, L82, O34. Keywords : Streaming, revenue allocation, multi-issue claims problems, pro-rata, user-centric. 1 Introduction The problem of adjudicating conflicting claims (in short, claims problem) models a basic situation in which an endowment is allocated among agents who have claims on it, and the available amount is not enough to fully honor all claims. This is a classic problem that can be traced back to ancient sources, such as Aristotle and the Talmud, although its formal treatment is somewhat recent (e.g., O’Neill, 1982). A sizable literature at the intersection of economics and operations research emerged afterwards (see, for instance, Thomson (2019) and the literature cited therein). Some of the efforts within that literature have been devoted to generalize claims problems to account for multiple issues (e.g., Calleja et al., 2005; Ju et al., 2007). 1 1 1 Acosta et al., (2022, 2023) have recently extended this multidimensional setting to allow for crossed claims. From a different vantage point, there has been a growing interest in recent years to analyze platform businesses, which have transformed the ways in which cultural content is produced and consumed (e.g., Aguiar et al., 2024). This is particularly the case with music. According to Statista, in the third quarter of 2024, Spotify (the largest music streaming platform) reached an all-time high with 640 million active users worldwide. The problem of sharing the revenue raised from paid subscriptions to streaming platforms among artists, a new form of revenue sharing problems under bundled pricing (e.g., Adams and Yellen, 1976; Ginsburgh and Zang, 2003; Bergantiños and Moreno-Ternero, 2015, 2016) is of critical importance to understand the management of music platforms. We show in this paper that such a problem can actually be approached resorting to the literature on (multi-issue) claims problems mentioned above. Traditionally, music platforms used a pro-rata method to award artists based on the relative number of total streamings they achieved in the platform. Gradually, they have been moving to a decentralized method in which the amount paid by each user is allocated among the artists that user streamed. If the latter allocation is proportional to individual streamings, we obtain the so-called user-centric method. If, instead, the allocation does not distinguish among artists that were streamed, we obtain the so-called Shapley method. Varying the allocation method, we obtain the family of probabilistic methods (e.g., Bergantiños and Moreno-Ternero, 2025). And compromising between the pro-rata and user-centric methods we obtain the family of weighted methods (e.g., Bergantiños and Moreno-Ternero, 2025). We show in this paper that all the methods described above can actually be derived from existing rules to solve (multi-issue) claims problems. Key among the latter are the so-called two-stage rules, which first divide the endowment among issues, and then divide the amount assigned to each issue among the agents. More precisely, we show that the family of probabilistic indices induces the same rewards as the family of two-stage rules where we use the classical constrained equal awards rule in the first stage, and any claims rule (satisfying two basic requirements) for the second stage. Similarly, we show that the family of user-weighted indices induces the same rewards as the family of two-stage rules where we use the classical proportional rule in the second stage, and any claims rule (satisfying a basic requirement) for the first stage. Our work can thus be interpreted as providing extension operators to convert rules to solve claims problems into rules to solve streaming problems. As such, we align with the literature dealing with operators on the space of claims rules (e.g., Thomson and Yeh, 2008; Hougaard et al., 2012; 2013a; 2013b; Moreno-Ternero and Vidal-Puga, 2021). We also align with the literature that applies the well-established framework from claims problems to solve various related allocation problems with real-life implications. Recent instances are water conflicts (e.g., Mianabadi et al., 2014), linguistic subsidies (e.g., Ginsburgh and Moreno-Ternero, 2018), broadcasting problems (e.g., Bergantiños and Moreno-Ternero, 2020), CO_{2} emissions (e.g., Duro et al., 2020), slots allocation in sport competitions (e.g., Krumer and Moreno-Ternero, 2020), or pollution abatement (e.g., Ju et al., 2021). The rest of the paper is organized as follows. In Section 2, we introduce the basic aspects of claims problems. In Section 3, we introduce the basic aspects of streaming problems. In Section 4, we provide the bridge between both types of problems, as well as our main results. We conclude in Section 5. 2 Claims problems Claims problems refer to an amount of a homogeneous and infinitely divisible good (e.g., money) to be divided among a set of agents, who have claims on the good that cannot be fully honored. See, for instance, Thomson (2019) for an excellent recent survey on the topic. A well-known example is bankruptcy. But many others can also fit this framework. Formally, a claims problem is a triple \left(N,c,E\right) where N is the set of agents, c\in\mathbb{R}_{+}^{N} is the claims vector and E\in\mathbb{R}_{+} is the amount to be divided. It is assumed that \sum\limits_{i\in N}c_{i}\geq E . Let \mathcal{C} denote the set of claims problems. A (claims) rule is a function R assigning to each claims problem \left(N,c,E\right)\in\mathcal{C} a vector R\left(N,c,E\right)\in\mathbb{R}^{N} such that \sum\limits_{i\in N}R_{i}\left(N,c,E\right)=E. Some examples of well-known rules are presented next. The constrained equal awards rule equalizes the amount received by each agent as much as possible, subject to the constraint that no agent gets an amount above his claim. Formally, for each (N,c,E)\in\mathcal{C} , and each i\in N , CEA_{i}(N,c,E)=\min\{\lambda,c_{i}\}, where \lambda satisfies \sum\limits_{i\in N}\min\{\lambda,c_{i}\}=E . The proportional rule yields awards proportionally to claims. Formally, for each (N,c,E)\in\mathcal{C} , and each i\in N , P_{i}(N,c,E)=\frac{c_{i}}{\sum\limits_{j\in N}c_{j}}\,E.\text{ } The weighted proportional rule yields awards proportionally to weighted claims. Formally, given weights \left(w_{i}\right)_{i\in N} , for each (N,c,E)\in\mathcal{C} and each i\in N, P_{i}^{w}(N,c,E)=\frac{w_{i}c_{i}}{\sum\limits_{j\in N}w_{j}c_{j}}\,E.\text{ } A multi-issue claims problem (e.g., Ju et al., 2007) is a tuple \left(N,K,c,E\right) where N is a set of agents, K is a set of issues, c=(c_{ij})_{i\in N,j\in K}\ where for all i\in N and j\in K, c_{ij}\geq 0 denotes the characteristic of agent i on issue j , and E\in\mathbb{R}_{+} is the amount of a homogeneous and infinitely divisible good to be divided. 2 2 2 Calleja et al. (2005) introduce multi-issue allocation situations , which are a particular case of multi-issue claims problems. Let \mathcal{MC} denote the set of multi-issue claims problems. For each \left(N,K,c,E\right)\in\mathcal{MC} and each j\in K , let c_{.j}=(c_{ij})_{i\in N} and C^{j}=\sum\limits_{i\in N}c_{ij}. A (multi-issue claims) rule is a function R assigning to each multi-issue claims problem \left(N,K,c,E\right)\in\mathcal{MC} a vector R\left(N,K,c,E\right)\in\mathbb{R}^{N}. A

Read the complete text