Mergers in network industries

BRIEFING NOTE
Mergers in Network Industries

Source
Ennis, S. (2003) "Mergers in Network Industries", Communications & Strategies, 50(2): 51-68.

Executive Summary
This paper analyzes the economics of network interconnection , focusing on point-to-point networks . The author explores how the relative size of networks and how users value the addition of new members (the value function ) influence the incentives to interconnect and the determination of interconnection fees, even in the absence of regulation. In addition, the paper examines the impact of network mergers on the value of interconnection for both merged entities and third-party networks, highlighting the crucial role of increasing or diminishing marginal returns in understanding these dynamics. The goal is to decipher the fundamental economic forces at work in interconnected network industries.
Key words
Network interconnection value
Unregulated interconnection charges
Impact of mergers on payments
Network value function
Marginal returns
Mergers and Interconnection in Network Industries

Main themes
•
Value of Interconnection in Point-to-Point Networks: The paper explores the incentives and value of bilateral interconnection between networks, focusing on how network size and the value function of members affect these dynamics.
•
Determinants of unregulated interconnection charges: The author examines why, in the absence of regulation, asymmetric interconnection charges may emerge, particularly in relation to network size.
•
Impact of Mergers on Interconnection Payments: The paper analyzes how mergers between interconnected networks can change the value of interconnection for both merged and non-merged networks, depending on the concavity or convexity of the network value function for consumers.

1. Nature of point-to-point networks and value of interconnection:

In a peer-to-peer network, the value to a member depends on the number of other members it can reach. Interconnection with other networks increases the quality of the product offered and potentially the price.

Interconnection makes a network both a supplier (access to its members for other networks) and a buyer (access to members of other networks).

The value of interconnection for a network is defined as the difference between its value when it is connected and its value when it is not.

Even with symmetric traffic flows, two networks may value their interconnection differently due to their respective sizes.
"When connected, the members of network i thus derive a value of interconnection with network j of Vi=niv(ni+nj)-niv(ni). Similarly, network j derives a value of interconnection with network i of Vj=njv(nj+ni)-njv(nj)."

2. Interconnection incentives and associated costs:

The incentive to interconnect arises from the increased value of the network to its members through access to a greater number of users.

When one party derives more value from the interconnection than the other, a negotiation based on NASH's bargaining solution leads to a sharing of the surplus, where the network that benefits more can pay the other.

In the case of networks of the same size, one might expect the absence of interconnection fees, which is consistent with practices between internet service providers of similar size.

The article finds an asymmetry in internet interconnection fees, where large networks generally do not charge each other but charge smaller networks. The model developed explains this asymmetry even in the absence of cost differences.

3. Role of the network value function for consumers:

The value function v(n) represents the value a user derives from belonging to a network of n users.

The concavity or convexity of this function is more critical than the strength of the network effect itself (measured by the slope v'(n) ).

Diminishing marginal returns (concave value function): The small network will value interconnection more than the large network. In this case, the small network can be expected to pay the large network for interconnection.
"Generally, when consumers experience decreasing marginal value from adding new users to their network, the benefit of interconnection is likely to be greater for the smaller network than for the larger network."

Increasing marginal returns (convex value function): The large network will value interconnection more than the small network. In this case, the large network can be expected to pay the small network for interconnection.
"Conversely, when consumers experience increasing marginal value from adding new users to their network, the total benefit to a large network of interconnection to a small network is greater than the value to the small network."

4. Implications of mergers on interconnection:

The impact of mergers on the value of interconnection depends strongly on the nature of the marginal returns in the value function.

Under diminishing marginal returns (v''(n) < 0): A merger between two networks (j and k) increases the value of interconnection for a non-merged network (i) relative to the merged networks. This implies that the non-merged network will pay more to the merged entities or receive less from them after the merger, because the loss in network size in the case of no agreement is greater.
"if v''(n)<0, the merger results in an increase in the value of interconnection to network i relative to the merged networks."

In case of increasing marginal returns (v''(n) > 0): A merger between two networks (j and k) decreases the value of interconnection for a non-merged network (i) relative to the merged networks. This suggests that the non-merged network might pay less to the merged entities after the merger.
"If v''(n)>0, the merger will result in a decrease in the value of interconnection to network i relative to the merged networks."

From the perspective of a non-merged network, the number of inaccessible users without a contract with the merged entity is greater than without contracts with the merging companies individually.

5. Sectoral points:

The hierarchical structure of the Internet (small networks purchasing services from medium-sized networks, which purchase from large networks) is consistent with an environment of diminishing marginal returns. Small networks can form coalitions to obtain more advantageous interconnection terms with large networks.

The existence of interconnection fees does not necessarily prove excessive market power, as there are real complexity costs associated with interconnecting with a large number of providers.

Examples of diminishing marginal returns include peer-to-peer file-sharing networks where new peers add fewer and fewer new unique resources.

Examples of increasing marginal returns might exist when the last critical connection to a network is made (e.g., the last desired contact joining a messaging system).

Conclusion

The article highlights the importance of network size and the shape of the consumer value function for understanding interconnection dynamics and the potential impact of mergers in network industries. In particular, the concavity or convexity of the value function determines whether small networks pay large ones for interconnection or vice versa, and how mergers affect the bargaining position of merged and non-merged networks. If the internet industry operates in an environment of diminishing marginal returns, a merger of large providers could lead to higher prices for paying customers and a conversion of non-paying customers to paying customers due to the increased size disparity.


Paper Summary Initial Draft By NotebookLM