Doubly-Bound Relationship Between Publisher and Retailer: The Curious Mix of Wholesale and Agency Models

Abstract

Since the time the industry for e-books transitioned from the wholesale to the agency model, several interesting observations have been made. For instance, despite its revenue-sharing approach that is supposed to be effective in addressing channel coordination issues present in wholesaling, the agency model has curiously led to higher prices and less demand in the e-book market. The main purpose of the current study is to provide insights into this curious market phenomenon by focusing on: (i) the substitutability between the printed- and e-book versions, and (ii) the interaction between the wholesale model (for printed books) and the agency model (for e-books). In this exploration, using a game-theoretic model, we discover that an increase in the agency fee that the retailer collects in the e-book market can depress the vitality of the market, which not only ends up hurting the publisher, but also the retailer. This is a telltale sign that retailers should perhaps be moderate in their effort to get a bigger cut in the agency arrangement. In addition, the complex interplay between the two parties across two retail channels is explained along with how it may lead to a counterintuitive equilibrium outcome. Some of the interesting observations from practice may, in fact, stem from a large agency fee in the e-book market. Reducing this fee can lead to a win-win outcome for both the publisher and the retailer.

Key words and phrases:

• Agency
• online markets
• books
• e-books
• economic theory
• retailing
• revenue sharing
• product substitutes
• wholesale markets

Acknowledgments

An earlier conference version of this work is: Kim, A. When Old Meets New: Wholesale and Agency Models in the Market for Printed and Electronic Books. In T. Bui (ed.), Proceedings of the 51st Hawaii International Conference on Systems Science, 2018, pp. 5174–5184. The author gratefully acknowledges the valuable feedback received from the referees and participants of the conference, particularly the co-chairs of the Strategy, Information, Technology, Economics, and Society (SITES) mini-track, Thomas Weber, Rob Kauffman, Rajiv Dewan and Eric Clemons, and the Editorial Assistant, Ai-Phuong Hoang. The author also appreciates the constructive feedback provided by the editors and reviewers of the Journal of Management Information Systems; the paper has greatly benefited from their efforts. Last, but not the least, the author is grateful to Debabrata Dey and Atanu Lahiri, not only for generously offering their wisdom during the development of this manuscript, but also for their constant moral support.

Notes

1. This demand specification comes from the following reservation price: $U\left(q_b,q_e\right)=q_b+q_e-\frac{\left({q_b}^2+{q_e}^2+2\sigma q_b{q}_e\right)}{2}$, where $\sigma =-\frac{{\mathrm{\partial }}^{2}U}{\mathrm{\partial }{q}_b{q}_e}$ represents the decline in the willingness-to-pay for one product arising from consumption of the other substitutable product [1].

2. In reality, the printed version may involve a non-zero marginal cost. If we were to consider this printing cost, $c$, in the model, the profit function for the publisher would change to: ${\mathrm{\Pi }}_P=q_b\left(w_b-c\right)+\left(1-\alpha \right)q_e{p}_e$. Following much of the same mechanics described in the paper, we can get the following equilibrium prices:

$w_b^{\ast }=\frac{2c\left(4\left(1-\alpha \right)+{\sigma }^2{\left(1+\alpha \right)}^2\right)+\left(1-\alpha \right)\left(8\left(1-\alpha \sigma \right)+{\sigma }^3{\left(1+\alpha \right)}^2\right)}{16\left(1-\alpha \right)+2{\sigma }^2{\left(1+\alpha \right)}^2},$

$p_b^{\ast }=\frac{4\left(3+c\right)\left(1-\alpha \right)-4\sigma \left(1-\alpha \right)+2{\sigma }^2{\left(1+\alpha \right)}^2-{\sigma }^3{\left(1+\alpha \right)}^2}{16\left(1-\alpha \right)+2{\sigma }^2{\left(1+\alpha \right)}^2},\mathrm{a}\mathrm{n}\mathrm{d}$

$p_e^{\ast }=\frac{8\left(1-\alpha \right)+2\sigma \left(1-c\right)\left(1+\alpha \right)-{\sigma }^2\left(1-{\alpha }^2\right)}{16\left(1-\alpha \right)+2{\sigma }^2{\left(1+\alpha \right)}^2}.$

As a sanity check, it is easy to see that all expressions match their counterparts when $c=0$. Once we repeat our analyses with these new expressions, we find that introducing $c$ in the model does not change our results qualitatively, although analytical tractability becomes an issue in performing comparative statics. Numerically though, we could verify that all the results remain applicable.

3. For an ordinary wholesale model with linear demand, the equilibrium wholesale price is $\frac{1}{2}$ and the corresponding retail price is $\frac{3}{4}$ cf. [36, p. 175].

Additional information

Notes on contributors

Antino Kim

Antino Kim ([email protected]; corresponding author) is an assistant professor of information systems at the Kelley School of Business, Indiana University. He earned his Ph.D. in information systems from the Foster School of Business at the University of Washington, Seattle. Dr. Kim’s research interests include supply chain of information goods, digital piracy and policy implications, misinformation and social media, and IT and worker displacement. His recent work is to appear in Management Science and MIS Quarterly. He has also presented his research at conferences and workshops such as INFORMS Conference on Information Systems and Technology, the Hawaii International Conference on System Sciences, and the Workshop on Information Systems and Economics.