Skip to Content
[CAIDA - Center for Applied Internet Data Analysis logo]
Center for Applied Internet Data Analysis
www.caida.org > workshops : wie : 1712 : abstracts.xml
Talk Abstracts: The 8th Workshop on Internet Economics (WIE 2017)
This page contains names, talk abstracts (if presenting), and topic keywords for workshop participants as they were submitted. Participants are encouraged to read these ahead of time to anticipate workshop discussion.
|   Workshop home    Agenda    Participants    Talk Abstracts    Exit Survey    Final Report   |

Participant Abstracts

NameAbstract
kc claffy (CAIDA)Talk Title: Introduction

Talk Abstract: kc or dcc will give the kickoff intro

Andrew Odlyzko (University of Minnesota)Talk Title: Some puzzles of Internet economics

Talk Abstract: Internet economics presents many interesting puzzles. For example, while most discussions of this topic concentrate on the high capital intensity of telecom, that is manifestly false. This industry is less about heavy capital investment, and far more about service, customer inertia, and territorial strategic plays (where the territories may be physical or virtual). That suggests different ways of thinking about communications business and communications policy

Interested in Discussing: What are the real economics of Google Fiber or some of the other competitive fiber providers?

Jason Livingood (Comcast)Talk Title: Wish Upon a Data Star: Data I wish I had, data I wish I could share, data I wish people talked about

Talk Abstract:Data I wish I had (which could help us improve the end user QoE)

  • Programmatic access to ingest edge provider OTT QoE data -- could be leveraged to improve access network performance and identify early signs of performance issues before they become user-noticeable
  • Detailed in-home WiFi performance data from COAM gear -- could be leveraged to improve the real-world end user experience since WiFi is the primary bottleneck
  • Quality & performance data from edge services being routed variously between direct interconnection links and 3rd party transit links (out of their control -- being done by their service provider) to quantify and understand the impact of interconnection on smaller edge services
Data I wish I could share
  • Detailed, per-network and per-location interconnect utilization data
  • Netflow-derived interconnect data that illustrates the regularity with which directly interconnected parties route traffic from under-utilized direct links to over-utilized or at-capacity 3rd party links (e.g. shining a light on peering playbook tactics as they are observed)
Data I wish people talked about
  • Data concerning platform concentration and how we classify someone as "essential infrastructure" as Cloudflare recently did "touching on the growing role of hypergiant edge providers, CDNs, and platform providers" and exploring the potential added fragility & dependence this introduces into the network over time (e.g. Dyn, AWS, Cloudflare). In addition, how do you even define and platform so that we can understand it and theorize about it?
  • Changes in return path routing of interconnected parties and the resulting changes in interface utilization, network performance, and QoE
  • The quantified impact of return path routing changes on smaller edge providers as well as any OTT service
  • Latency under load being a critical aspect of end user QoE -- not just throughput
  • Relative IPv6 vs. IPv4 performance & native dual stack vs. Carrier Grade NAT (CGN) -- impacting end user QoE

Johannes Bauer (Michigan State University)Talk Title: Drivers of innovation in the advanced communications ecosystem

Talk Abstract: Digital innovation can be modeled as an evolutionary process of trial and error. Successful innovations are selected in a market or non-market environment and replicated. Several types of complementary innovation processes unfold in parallel and are shaped by different sets of factors. Some innovations work well in a framework of modular technology, others require different forms of market and non-market coordination. If such mechanisms do not exist, innovations may not take place at all, they may be slowed down or only be offered on a sub-optimally limited scale or they may migrate to closed networks. The paper discusses the role of such coordination mechanisms and how we can assess the rate and direction of different types of innovation in the Internet ecosystem. It argues that modular edge innovation is only one type and that other types may become more relevant in the future. If this is correct, new types of coordination deficiencies and failures may shape the path of innovation in wanted and unwanted forms. The paper concludes with some thoughts in policy implications.

Interested in Discussing: 2b, 2c, 3c

William Lehr (Massachusetts Institute of Technology)Talk Title: Regulatory Governance and Policy for the Digital Economy

Talk Abstract: Building on our paper "Communications Act 2021" (with Doug Sicker), one of the key issues warranting further investigation is the need for a communications sector specific regulator (e.g., the FCC), the appropriate scope for its responsibilities, and how that compares with devolving the FCC's responsibilities to other regulatory bodies or agencies.

Interested in Discussing: Cybersecurity risk measurement strategies (a close parallel to reliability measurement)

Costas Courcoubetis (SUTD)Talk Title: Policy issues in ride sharing

Talk Abstract: It is widely believed that ride sharing, the practice of sharing a car such that more than one person travels in the car during a journey, has the potential to significantly reduce congestion by filling up cars more efficiently. By analyzing a game-theoretic model will large number of players we find that the ratio of ownership to usage costs is the crucial determinant that decides how ride sharing is organized in society. If this ratio is low, ride sharing is offered as a peer-to-peer (P2P) service, and if this ratio is high, ride sharing is offered as a business-to-customer (B2C) service. In the P2P case, rides are initiated by drivers only when the drivers need to fulfill their own transportation requirements. In the B2C case, cars are driven all the time by full time drivers taking rides even if these are not motivated by their private needs. Our analysis specifies how car ownership and congestion are effected by platform prices, cost parameters, and the distribution of the individuals' types. In particular we discover that traffic and ownership may increase as the ownership cost increases, and that a revenue maximizing platform might prefer a situation where cars are driven with only a few seats occupied, creating high congestion. In this talk we like to discuss regulation policy design issues, avoiding excess congestion but profiting from the sharing economy.

Interested in Discussing: What should a new Telecommunications Act look like?

Emanuele Giovannetti (Anglia Ruskin University)Talk Title: Exploring the links between Networks' hierarchical structuring and affordability. Initial results from Tamil nadu

Interested in Discussing: How to develop policies to address/ impact any emerging relation between Internet hierarchical network structuring and affordability. Specific interest on these policies in developing countries.

Scott Wallsten (Technology Policy Institute)

Interested in Discussing: I have studied willingness to pay for bandwidth and latency, universal service issues, auctions, net neutrality, political economy of telecom regulation and legislation. Happy to talk about any, or none!

Scott Jordan (University of California, Irvine)Talk Title: The FCC regulates services, not networks

Talk Abstract: Discussants at WIE often confuse regulation of services with regulation of networks. In this talk, I'll provide an overview of the relationship between services and networks based on a history of FCC Orders. I'll give examples of services that have been offered over the PSTN and over the Internet. I'll also give examples of different types of regulation of services offered over the same network.

Interested in Discussing: Given that a single network can be used to offer multiple services, how can we determine when a service should be regulated, and what type of regulation is appropriate?

James Miller (FCC / OET)Talk Title: Challenges and a Path Forward for Realizing Data-Driven Federal Internet Policy

Talk Abstract: A significant challenge for policy makers is the baseline facts for a dispassionate understanding of the key points impacting the Internet economy. At the same time that access to basic datasets related to consumer broadband pricing remains a significant problem for basic policy analysis and progress has been made in the development of datasets looking at Internet performance, basic programs for deployment and adoption are undergoing challenges. Access to data is an area of concern and crowdsourcing and other collaborative efforts offer significant opportunities. The FCC remains one of the top "big data" agencies in the volume of data collections permitted by OMB. Nevertheless there remain limited staff and computing resources to make data usable in policy decision-making. Strategies to motivate and train staff to use modern tools represent a hidden gem that could help meet the needs to data analysis. The FCC another other Internet economy agencies also are undergoing a transition from the sole active parties in these topics to participants and in some cases spectators in the resolution of many debates. These agencies must better accomodate the sharing and presenting of data to the public and fullfuill a fuller role as a facilitator in data-driven policy. Both staff and computing resources need to better prepare to meet this goal but also developing new and existing collaborative relationships is a possible key to success as a the enabler of a multi-stakeholder driven data and evidence-based process.

Interested in Discussing: I'd like to focus some discussion of the current state and path forward to addressing the role of data at the FCC including a discussion of impacts of bad data policy on the current state and unrecognized potentials for better path forward.

Christopher Yoo (University of Pennsylvania)Talk Title: Product Life Cycle Theory and the Maturation of the Internet

Talk Abstract: I propose to explore the implications of saturation of U.S. wireline broadband market by examining the literature on the product life cycle, dominant designs, technological trajectories and design hierarchies, and the role of complementary assets in determining industry structure. These theories suggest that the focus of competition will shift to process improvements to delivering more differentiated, higher value services.

Vishal Misra (Columbia University/Google)Talk Title: Differntial pricing and Network Neutrality

Talk Abstract: In this talk we explore the connection between the notion of Network Neutrality and Differential Pricing. Specifically, we will investigate the impact of schemes like Zero Rating on Network Neutrality and Competition.

Interested in Discussing: Should regulatory bodies make a ruling on differential pricing or view them on a case by case basis. For context, India and Canada have made a ruling, FCC prefers a case by case approach.

Marvin Sirbu (Carnegie Mellon University)

Interested in Discussing: The future of common carriage in a new Telecom Act

Erin Kenneally (Dept of Homeland Security)Talk Title: (to the extent relevant) Cyber Risk Economics Program at DHS CSD
Arnaud Jacquet (BT)Talk Title: structure and evolution of bradband demand

Talk Abstract: We've seen an explosion of broadband traffic since 2000. Can we understand better the impact of the different drivers of demand and QoE in a very dynamic broadband ecosystem (comms operators, device manufacturers, content providers, users)? What are the metrics that matter to the different stakeholders?

Interested in Discussing: how best to define broadband usage? (how do we know that a user's experience is adversely affected by lack of speed?)

Tony Tauber (Comcast)Talk Title: An Essential Part of the Internet

Talk Abstract: What components of the Internet are essential? How would/do we know? What measurements and research reveal them? What can/should/need be done about them, if anything? Can history help us think about the topic?

Richard Clarke (AT&T)Talk Title: Will FTTH be the principal technology for very high speed Internet access?

Talk Abstract: For years FTTH was considered to be the eventual solution for nearly all very high speed Internet access. But as time has progressed, its business case has remained problematic -- primarily due to improvements in the economics and capabilities of alternative technologies. Is FTTH still the eventual solution? Or has the age of "peak FTTH" passed?

Robert Frieden (Penn State University)Talk Title: The Internet of Platforms and Two-Sided Markets: Implications for Competition and Consumers

Talk Abstract: This paper examines developments in the Internet marketplace that favor embedded intermediaries that have significant market power and the ability to install a platform that both upstream sources of content and applications as well as downstream consumers need to access. Ventures such as Amazon, Facebook and Google have exploited, "winner take all" networking externalities resulting in the creation of seemingly impenetrable barriers to market entry even by innovative companies. Courts and regulatory agencies recognize the substantial market shares these ventures have acquired, but refrain from imposing sanctions on grounds that consumers accrue ample benefits when platform operators use upstream revenues to subsidize downstream services. Consumers also benefit when platform intermediaries eschew short term profits in the quest for greater product diversity and "shelf space" in the Internet marketplace. Additionally, courts and regulators may over-estimate the opportunities for consumers to migrate to alternatives, but underestimate the harms to competition and consumers occurring on either or both sides of an intermediary's platform. The paper identifies four types of government responses to price and quality of service discrimination that exploits choke points within the Internet ecosystem where large volumes of traffic has to traverse a single digital, broadband carrier's network, or service provider's platform. Governments can refrain from regulating access and accept aspects of market concentration as proper rewards to ventures offering desirable content and carriage services. Alternatively, they can impose access sanctions for antitrust violations, unfair trade practices, consumer harms and unreasonable access discrimination to offset market-driven concentration and dominance. Between these poles, governments can apply antitrust/competition policy remedies, or rely on expert regulatory agencies to respond to complaints. The paper examines digital broadband platform operators with an eye toward assessing the aggregate benefits and costs to both upstream firms and downstream consumers. The paper notes that in some instances, government-imposed, ex ante safeguards anticipate and attempt to resolve transitory, comparatively insignificant, or possibly nonexistent problems. In other instances, governments have stated an intention to rely on ex post remedies, but they never get around to refining procedures to reduce the potential for false negatives that ignore or underestimate significant marketplace harm. The paper concludes that governments have failed to revise and recalibrate tools that examine potential marketplace distortions and assess the potential for damage to competition and consumers. The article demonstrates how the Justice Department, Federal Trade Commission and the Federal Communications Commission have relied on economic and legal doctrine ill-suited for digital broadband market assessments. These agencies have generated false positives, resulting in market intervention where no major problem exists and false negatives, where undetected major problems cause harm without remedy. Additionally these agencies appear to misallocate their resources and attention on insignificant matters when more compelling problems exists. The paper recommends that courts and government agencies consider both sides of an intermediary's market when assessing current and prospective harms, commit to more accurate estimates of the potential for consumers to change platform allegiances and consider the overall impact of two-sided market domination instead of simply dismissing market concentration as necessary and harmless.

Mark Johnson (MCNC)Talk Title: non-traditional provider issues - policy and economic

Talk Abstract: I would like to describe the problem that outside of wealthy urban or suburban areas traditional providers don't see an opportunity to make their traditional ROI. Alternative providers including non-profits and municipalities have extra-economic (your economist friends probably have better terms) motivations for delivering proper broadband to these areas. It is also the case that these economically deprived areas don't generate enough business to have competition for service there (in spite of the usual rhetoric that competition will solve everything).

Interested in Discussing: How do we use data to drive policy improvements leading to more Internet

  Last Modified: Mon Jul-6-2020 22:31:08 UTC
  Page URL: https://www.caida.org/workshops/wie/1712/abstracts.xml