Abstract
In the 20 years since the start of the World Wide Web, the Internet has evolved in such a way that we can feel its impact on most aspects of our lives, including our connections with others. It provides for more interactive information exchanges than previous technologies permitted, and the speed at which information now travels expands personal worlds while shrinking the global society. These new technologies have great potential to advance society in more democratic directions, but there is no certainty that these potentials will be realized. This article provides an overview of some of the available policy alternatives in the network neutrality discussion. These alternatives attempt to address last mile Internet service providers’ treatment options of end-users’ data packets, both in terms of content-based data transfer decisions and cumulative data transfers. These alternatives are evaluated in terms of the effects they could have on the Internet’s ability to further democratic pursuits in the US and other Western democratic societies. Considering current capacity limitations of networks, I ultimately suggest a system of consumer-driven prioritization and a usage cap at 90% of all users with overage charges for the top 10%. This is not a perfect solution from a stance of furthering the democracy-encouraging potential of the Internet and existing technologies, but they appear to be the most realistic options that do not completely sacrifice democratic potential. It is my hope that through this article and similar efforts that a public deliberation of these policies’ merits and feasibility could begin.
Introduction
In the 20 years since the start of the World Wide Web, the Internet has evolved in such a way that we can feel its impact in many aspects of our lives, including our connections with others. It provides for a more interactive form of information exchange than previous technologies permitted and the speed at which information now travels expands our personal worlds while shrinking the global society. These new technologies have the potential for us to be more connected, more detached, more invaded, more democratic, more community-oriented, and more consumption driven.
A topic of continued presence in the news over the past several years has been what is known as “net neutrality.” Sandvig (2007), referencing Wu’s (2003) work on establishing a set of rules governing appropriate “network neutrality,” notes that network neutrality (or “net neutrality”) means that while a last-mile Internet service provider may prohibit the transmission of certain data packets because of bandwidth capacity issue concerns but, rather, that it may not discriminate against or in favor of certain packets of data traveling over its infrastructure on the basis of the content of the packet (p.138). The issue was integrated into both major candidates’ presidential campaigns in 2008, though that was not its first entrance into public awareness (Mark, 2007; Broache, 2007). Additionally, the Federal Communications Commission (FCC) has been actively engaging in dialogue from 2008 through 2011 with both Congress and the public to determine the correct framework with which to understand how the Internet should be regarded in terms of the agency’s regulatory authority and what role the government should play, if any, in the relationships between and among Internet service providers (ISPs), content providers, backbone supporters, and end-users. There have been high tensions with regards to this issue particularly between the FCC and ISPs, as was demonstrated in the United States Court of Appeals case, Comcast vs. FCC (2010), in which the District Court of Appeals stated that the issue of net neutrality is an important one to consider but the FCC does not have the authority or jurisdiction to regulate how Internet service providers treat information packets received from content providers under current Congressionally-derived authority and its classification of the Internet as an ‘information service’ subject to the least restrictive regulations under the 1996 Telecommunications Act.
My purpose here is to provide an overview of some of the available policy alternatives that address how so-called last mile Internet service providers could treat data packets flowing to the Internet end-user, both in terms of content-based data transfer decisions and the treatment of cumulative data transfers. These alternatives are evaluated in terms of the effects they could have on the ability to use the Internet to further democratic pursuits in US and other Western democratic societies. I argue that the Internet has the potential to improve democracy’s prospects but that this future is nowhere near assured, especially if private corporations are permitted to make determinations of which types of content are allowed to travel to end-users and at what speeds.
Roadmap
I begin this article by providing a clear context of the network neutrality discussion as an evolution of the digital divide issue. From there, I will elaborate on recent governmental-side actions in terms of the Federal Communications Commission’s National Broadband Plan and the Comcast v. FCC court decision. Following this contextual positioning will be a brief review of the literature regarding the Internet as a mechanism for democracy. Within this section, I will focus first on the application of Habermas’s criteria for a deliberative democracy on the potentials allowed by the Internet. For the purposes of this article, I will utilize the teleological standpoint of the Internet as a mechanism to further the cause of democracy in a society, as opposed to the alternative viewpoints of Internet-for-commerce and Internet-for-entertainment, mentioned above.
With this groundwork laid, I will delve into my central purpose: to review some of the major dimensions of policy alternatives I have observed in terms of who should control the type of content that flows to the Internet’s end-user (Content-Based Arrangements) and how companies could regulate the volume of traffic that travels over their Internet infrastructure (Cumulative Data Transfers). After indicating which interests have promoted the specific policy alternative and their likely impetus for doing so, I will offer an evaluation of how, if enacted, the alternative could promote or inhibit efforts to increase the democratic potential in a society.
I will wrap up the article with a couple of comments on what this discussion can mean even if the Federal Communications Commission is not formally granted the authority to regulate last-mile Internet service providers. As this is a constantly evolving field, material quickly becomes outdated as deliberations over appropriate regulations proceed and as technology evolves. I conclude by positing that a conversation on the possible policies for handling the flow of information on the Internet could actually begin a public deliberation and move us in the direction of citizens engaged in a Habermasian democratic society, virtual or otherwise, rather than a society of passive consumers who merely absorb that which is presented to them.
Context for the Discussion
The Digital Divide’s Evolution
The United States National Telecommunications and Information Administration (NTIA) describe the digital divide in terms of access to equipment and infrastructure. As a result, efforts to counteract the digital divide have focused on increasing broadband access and the numbers of computers in libraries and other public facilities (Stevenson, 2009, 4-5). Valadez & Duran (2007) argue that simplifying the digital divide to access negates “inequalities in technology and learning” that result in “vast differences in opportunity, experiences, and practices” (p.34). They argue that such a complex issue requires more complex understanding and solutions than throwing computers at schools and public libraries. Valadez & Duran (2007) discuss the three dimensions of access as motivation, possession, and skill:
Motivation refers to the willingness of individuals to use technology and to include it in their home, work, and educational efforts. Possession describes a more concrete definition of access including physical access to [computers and the Internet] and the ability to use the technology. Skills refer to the ability to use the technology, and the degree of support available to instruct individuals in its use (p.33).
Becker et al. (2010) make the claim that new digital divides emerge with the introduction of each new wave of technologies. As these new technologies penetrate society, those without access to even the older technologies fall farther and farther behind, even as the overall gap begins to close. The new divides include “available Internet bandwidth, quality of computer equipment, and the ability of users to successfully navigate the Internet to accomplish their goals” (Becker et al., 2010, p.15).
The discussion of network neutrality is a natural evolution of concerns surrounding the digital divide. Even after all Americans who desire to be connected to the Internet have been connected, their connections will matter little if the service providers deny meaningful access to certain types of content either through outright prohibition or through regulating the speed at which the material can be accessed relative to the speed of access of other sources of information. Additionally, just because areas and individuals have adequate access in terms of landlines or wireless coverage and speed of connection, these will only be short-term solutions if last-mile Internet service providers do not continually reinvest in the technologies to keep up with the increasingly bandwidth-intensive demands websites place on their infrastructure to deliver information to end-users.
National Broadband Plan
The American Recovery and Reinvestment Act of 2009 statute enacted by Congress included a provision for the development of a National Broadband Plan that would provide a roadmap for how to deploy broadband Internet service delivery to all Americans (Federal Communications Commission, March 2010). The Plan itself and the Connecting America supporting document was crafted by the Federal Communications Commission with the stated purpose of making the United States the leading technological force again. However, less than a month after the release of this comprehensive plan of action, the United States Court of Appeals released their ruling in the case of Comcast Communications v. Federal Communications Commission and United States (2010).
Comcast vs. FCC
In August of 2008, the FCC issued a cease and desist order for Comcast to end its practice of network management that regulated peer-to-peer file sharing information separately from other types of content transmitted by the service provider to end users (Federal Communications Commission, 2008, p.1). This followed a November 2007 complaint filed by several entities, including Free Press and Public Knowledge, to the FCC regarding this practice and the claim that such a practice was in violation of current regulatory and statutory codes. Comcast initially agreed to cease this practice but followed that agreement with a request for review by the Court of Appeals, alleging that the FCC had overstepped its regulatory bounds.
While the court case was pending, the FCC began a Notice for Public Rulemaking period, as specified under the Administrative Procedures Act, that would attempt to clarify the FCC’s authority and scope of regulating issues related to Internet and Communications Technologies (ICTs). The intended result of this rulemaking period was to codify a set of principles they proclaimed would “preserve the open Internet” (Federal Communications Commission, 2009). Four of the principles were already in practice and stated that consumers were entitled:
- To access lawful content of their choice;
- To run applications and use services of their choice, subject to the needs of law enforcement;
- To connect their choice of legal devices that do not harm the network; and
- To competition among network providers, applications and service providers, and content providers. (Federal Communications Commission, 2009, p.3)
In addition to these initial four principles, the FCC recommended two further principles be adopted and codified, which would state that Broadband internet service providers must:
- Treat lawful content, applications, and services in a nondiscriminatory manner; and
- “Provide reasonable levels of network management practices transparency in such a way that would ensure provider compliance with the other five stated principles” (Federal Communications Commission, 2009, p.5).
In April 2010, the United States Court of Appeals for the Fourth District ruled in Comcast Corporation v. Federal Communications Commission and United States that the FCC does not have the jurisdictional authority under the Telecommunications Act of 1996 to bar the Internet service provider, Comcast Communications, from prohibiting the use of peer-to-peer file transferring programs by its customers. At the same time, however, the Court of Appeals did reaffirm that the ruling in Comcast’s favor was on the basis of a lack of agency authority and not based on the principle in debate.
In evaluating previous FCC decisions of Internet service provider classification, the FCC essentially caused this problem themselves in 2005 when they attempted to reconcile conflicting regulations for cable Internet providers and DSL Internet providers. At the time, cable ISPs were treated as “information services” under Title I of the Telecommunications Act of 1996, a much less stringent regulatory environment than Title II, while DSL Internet providers were being regulated by the FCC as “telecommunications providers.” To correct this situation, the FCC reclassified DSL as an information service to bring it in line with the regulations for cable Internet service rather than moving cable Internet service providers’ regulations to be in line with DSL.
With the Court of Appeals ruling that the FCC did not have power delegated to it by Congress to make regulations and rulings regarding “open Internet,” the FCC opened a Notice and Comment period and held public hearings in 2010 to solicit suggestions and recommendations for a new way to conceptualize the Internet in order to best understand what the FCC’s role should be in regards to Internet-related regulations. Several of the policy alternatives presented in this article come out of the comments received from various interest groups responding to this request.
Internet-for-Democracy
In discussing how policies regarding Internet and service provision should be implemented, it is important to consider the perspective we are using to think about the Internet and its purpose. I have identified three main purposes of the Internet in the American context that influence potential policies: Internet-for-commerce, Internet-for-entertainment, and Internet-for-democracy. Certain policies may look favorable when they are viewed through the Internet-for-entertainment lens but harmful when regarded through the Internet-for-democracy perspective. For the purposes of this assessment, I will utilize the Internet-for-democracy perspective to evaluate policy alternatives. Using this perspective and even implementing policies identified as favorable to these ends are not guaranteed to bring forth democracy. Instead, enacting “Internet-for-democracy”-based policies simply improves the conditions in which a society conducive to democratic activities may exist. Dahlgren (2009) notes that “discussions about media and democracy are often framed by the notion of the public sphere, which emphasizes that the media must provide citizens with the information, ideas, and debates about current affairs so as to facilitate informed opinion and participation in democratic politics” (p.34).
After evaluating other authors’ perspectives regarding the relationship between the Internet and democratic society (McChesney, 2004; Dahlgren, 2009; Margolis & Moreno-Riaño, 2009; Sunstein, 2007; Morozov, 2011; Goldsmith & Wu, 2006), I have attempted a cyber-realist approach to the actual possibilities enabled by the Internet (Morozov, 2011, pp.318-320). From this perspective, cyber-realism acknowledges that the Internet holds no “silver bullet” for eliminating authoritarian regimes or ushering in new eras of democracy just by its very existence (p.319). While the Internet has improved the odds of our society becoming more democratic, ultimately, I view the Internet as being an extension of physical society. Just like our physical world, the Internet has the potential to promote democratic values by making it easier to connect with others over geo-spatial and temporal limitations. Just as easily, there is the potential that democratic values could be reduced either through the overt actions on the part of governments or businesses or through the embrace of our roles as consumers. This consumer-based focus would emphasize the Internet-for-commerce and Internet-for-entertainment purposes of the Internet over its democratic potential.
The Internet and Deliberative Democracy
According to Margolis and Moreno-Riaño (2009), “deliberative democracy… is a cooperative process of collective decision-making that is dynamic, inclusive, information-based, and, as some argue, vital for invigorating the waning democracies of the 21st century” (p. 52-53). Historically, the potential for realizing a democratic society founded on the principles of deliberation has been limited for all but the smallest societies. Geo-spatial and temporal limitations easily impeded the ability to deliberate in a large and diverse society. The Internet opens up the possibility of transcending these limitations to connect with people across vast distances and in an asynchronous manner in order to reduce some of the main external restrictions on participation. Technology can only work to reduce external deterrents from participation and is nearly powerless to counteract forces internal to an individual that would preclude them from participating (Froomkin, 2003, p.869).
Policy Alternatives for Consideration by Internet Service Providers
There are two main policy areas for consideration in discussing how data should be treated by last-mile service providers: how data packets are qualitatively treated by the last-mile service provider, referred to in this analysis as “Content-Based Arrangements;” and how cumulative data usage should be treated. Three network management approaches are apparent within the Content-Based Arrangements framework: common carrier, consumer-driven prioritization, and provider-directed prioritization. Additionally, I review four types of cumulative data transfer arrangements in this analysis: flat rate for unlimited usage, tiered data usage packages, pay-per-byte, and monthly data usage caps plus overage charges.
Two aspects of network performance that should be kept in mind during this discussion are application sensitivity and congestion. Applications are broadly classified as “time-sensitive” or “non-time-sensitive.” Time-sensitive applications are then ranked further on the basis of sensitivity within the category. In its report to the FCC regarding a new congestion management tool, Cox Communications (2010) provided examples of the types of applications and functions. Time-sensitive applications include:
- VoIP (Voice over IP, telephone calls made over the Internet);
- Games (Online interactive games);
- Streaming (Web-based audio and video programs);
- Web (Web surfing, including web-based email and chat embedded in web pages);
- IM (Instant messages);
- Tunneling and Remote Connectivity (VPN-type services for telecommuting);
- Email; and
- Other (Any service not already categorized). ( p.25)
Within this list, VoIP is likely to be the most sensitive to latency and packet loss in terms of the overall Quality of Experience (QoE). Data packets must be delivered in real-time and quickly in order to maintain voice clarity and connectivity. While it may be surprising that streaming does not rank higher in this list, the reasoning for its placement is that most streaming programs, including Netflix, Hulu, YouTube, and Pandora, “buffer” their content so that data packets are stored on your computer for several seconds before the application actually needs them in order to maintain a smooth video or audio experience.
Non-time sensitive applications, as categorized by Cox Communications (2010) include the following:
- File Access (Bulk transfers of data such as FTP);
- Network Storage (Bulk transfers of data for storage);
- P2P (Peer to peer protocols);
- Software Updates (Managed updates such as operating system updates); and
- Usenet (Newsgroup related). ( p.25)
The delivery of data related to these applications can be deprioritized in relation to the data streams of time-sensitive applications with little to no impact on a user’s overall quality of experience.
Congestion is a general term used to describe the amount of data traveling over a network connection in relation to the total amount of data that a network can support traveling over it at a given time. Network infrastructure has a finite carrying capacity and end users’ experiences will likely be negatively affected as the amount of traffic over the connection increases. Some of the proposed network management policies would be in effect at all times on a network while others would only be used when the load on a network reaches a certain percentage of the overall capacity.
Content-Based Arrangements
Common Carrier
The term “common carrier” dates back to English common law and indicates that the entity in question shall provide service to all who wish to pay for its services and will treat those individuals and goods equally. In terms of telecommunications, this means that if an individual pays for telecommunications service, in this case meaning telephone service, then the telecommunications corporation is required to connect all calls being sent by the user and permissible within the terms of the contract and all calls being received by the user from other telephone networks. The provider is prohibited from favoring one type of call over another type of call as long as the user is paying for the service. In regards to the Internet, classifying Internet service providers as common carriers would hold them to the same standard and require that they treat all types of data transmitted and received equally (Sandvig, 2007).
Classifying Internet service provision as an information service is logical when those who are providing the service are not the same as those who maintain the infrastructure itself. In the days of dial-up Internet service, one’s service provider often differed from one’s telephone service provider. Now, companies like Comcast, Cox, Verizon, AT&T, and Time Warner are filling two roles: they are providing both the access connection and the related services. To regulate ISPs under Title II would force them to operate the access connection under common carrier guidelines with non-discrimination practices towards both content and connections. Such a consideration would meet net neutrality advocates’ request for neutrality on the part of service providers because ISPs would be transmitting and receiving all data packets the same as any other they receive in terms of both bandwidth allocation/usage and latency. Service providers are aware that we primarily use them for the connection to the broader network and related speed and not the “extras.” This understanding on the part of service providers is obvious from how they advertise their services. For example, Comcast Communications advertises “incredibly fast Internet speeds” that allow one to:
- Watch streaming HD movies online.
- Play games online with less lag.
- Download music and upload photos in the blink of an eye with PowerBoost
- Get faster Internet connections-even with the whole family online at the same time. (Comcast website)
The benefits of this approach to determining appropriate treatment of content by ISPs include that it does not allow for ISP-driven censorship, a benefit that is celebrated by many members of social movements who fear having their voices silenced on the Internet because their views do not correspond with those of the ISP. Advocates argue that a “data packet is a data packet,” no matter the content. It does not matter if the information contained in the packet will be reassembled into a large movie file or a small Word document because the packets can flow the same with either format. The main difference is how many packets comprise a file. From this perspective, ISPs should serve as common carriers and should be mandated to transmit all legal content equally over their networks.
Opponents to this approach argue that consumers’ current data transfer levels are increasing rapidly as more and more activities are taking place online and the activities that are taking place are largely more bandwidth-intensive than in the past. They claim that the increasing rates are simply unsustainable, even with massive capital investments to improve provider infrastructure. Opponents state that without a decision that breaks away from network neutrality advocates’ model of “all packets are equal,” network congestion will continue to increase to unmanageable levels and end-users will likely find their Quality of Experience degraded from this congestion.
Consumer-Driven Prioritization
Keeping in mind complaints by ISPs about the common carrier model of content treatment, the next policy alternative for consideration is prioritization of content based on consumer/user generated preferences. This model relies on a perspective that not all packets are equal in terms of their importance to the overall QoE for the consumer and that the consumer is best suited to prioritizing packet delivery based on their predominant ways of using the Internet. During business hours, a teleworker may want to prioritize VoIP for calls with employers or clients and VPN connectivity for work activities. In the early evening, streaming may take priority out of a desire to stream movies from Netflix or other sites. Late at night, the same end user may want to play World of Warcraft and log on with some friends. This perspective emphasizes that it is the user who is best able to determine his or her preferences. Cox Communications and Sandvine are both influential technology companies that support this approach, with the former being an Internet service provider and the latter offering technical support and assistance to Internet service providers. In its comments to the FCC, the Center for Democracy and Technology (2010) also presented “user-directed prioritization” as an acceptable network management model.
An important consideration in this approach is determining how well an application utilizing the network can handle latency, which is the amount of time it takes for packets to reach the end user. Consumers input their priorities from a list of application and/or data types that reflect the common ways that individuals use the Internet, including streaming, online gaming, file transfers, and VoIP. One possibility is that the service provider, after analyzing common sources and destinations of content, would set default prioritization preferences that the user could change if their preferences deviated from the norm in order to improve their experience.
In light of democratic-potential concerns, we should approach this method cautiously. Margolis and Moreno-Riaño (2009) call attention to the “citizen is consumer” syndrome that threatens to conflate the private desires of consumers with those actions and preferences that would reflect the responsibilities of citizens to the larger society and the need to be an informed citizenry that can properly participate in a deliberative society (pp.-65). The personal freedom to choose what to consume is merely one dimension of a democracy and we should take care to keep such limitations in mind as we evaluate potential alternatives. “Citizens are not supposed merely to press their own self-interest, narrowly conceived, nor are they to insulate themselves from the judgment of others” (Sunstein, 2007, p.37). Sunstein (2007) also emphasizes the danger of information personalization, what he refers to as the “Daily Me,” because it can lead to balkanization and radicalization within groups (p.63). If individuals are no longer encountering dissenting views from their own or information on topics that do not strongly impassion them, then they are likely not being exposed to general-interest or news sources that would expand their perspectives and potentially make them more amenable to deliberation than a silo-ed society in which they are only exposed to opinions that are like their own (Sunstein, 2007, p.48).
Provider-Directed Prioritization
The service provider industry refers to provider-directed prioritization as “traffic shaping.” In this approach to network management, Internet service providers determines which kind of packets should flow through its network connections at what speeds to whom. This decision may be based on monetary considerations, how the application may affect end users’ Quality of Experience, end-users’ network utilization rates, and/or network congestion levels.
Specifically, the ability of content providers to pay for prioritization of content is one of the main concerns about ISPs having the freedom to regulate their networks on the basis of data packet contents or source. Viewing this technique through the lens of Internet-for-democracy illuminates that this approach will have a negative impact on a society’s democratic potential. In the best case scenario, providers will prioritize the content they think consumers want (Cox Communications, 2010). Another potential scenario that also has a poor outlook for democratic potential in societies is for an ISP to prioritize its own content and/or content paid for by affiliates such that either there is a fast stream of owned or affiliated content flowing to end-users or that non-affiliated content will be slow through either congestion of smaller pipes for such information or through a conscious effort to throttle, or “slow down” certain types of information flowing over a network’s infrastructure. The worst imaginable provider action within this particular framework is for people in positions of power to eliminate all access to potentially objectionable/questionable content, even including threats to individual ISP’s market position. This perspective is mostly associated with actions on the part of illegitimate governments seeking to control aspects of society through a triangle of communications action: surveillance, propaganda, censorship (Morozov, 2011).
Most opponents to this perspective concede that some degree of network management is required but still express concerns that “application and content providers unwilling or unable to pay the required ‘tribute’ could be cut off altogether from the last-mile provider’s end users, or otherwise be forced to accept a degraded connection” (Selwyn & Golding, 2010, p.123).
Pipe Segregation
One of the proposed nuances that has generated simultaneously more and less outrage from many in the net neutrality movement is the possibility of instituting enhanced services priority pipes within a network. This concept is at the core of the Verizon-Google legislative proposal released in August 2010, which advocates a core policy of nondiscrimination against content but allows Internet service providers to provide faster access to some enhanced services (Verizon & Google, 2010).
Opposition has come from many organizations within the SavetheInternet coalition, led by Free Press, while organizations such as the Center for Democracy and Technology have indicated that they would support such a proposal if the fast-track pipes would not be allowed to expand to the entirety of the network connections and as long as the FCC has stronger regulatory authority regarding the general access to the Internet in order to ensure that service providers are not crowding out non-prioritized traffic.
One suggestion that has some degree of merit, though still with a great deal of caution, is for pipes to be one-way permeable, allowing general access traffic to use bandwidth allocated for enhanced services in the event that enhanced services have excess bandwidth but making sure that there is a minimum threshold for general access, such as 60-70% of total bandwidth, and that enhanced services could not utilize that portion of the bandwidth, even if general access service usage was not at capacity. This is actually a similar approach to that of the Post Office, in which the standard mail delivery setting is First-Class, but if certain items are deemed to have higher priority, customers can select Express or Next-Day mail to have faster delivery. The existence of Express mail does not prohibit the transport of First-Class mail, but First-Class mail may be delivered faster in situations in which there is low demand for Express mail services. Other pipe segregation options include strictly segregated pipes in which there can be no spillover from one category to another regardless of pipe capacity and a completely permeable pipe in which general access and enhanced services share the same section of bandwidth but enhanced services travel through the connection faster than general access-oriented packets.
Data Usage Regulation by Types of User
Some ISPs, including Comcast Communications, have been known to deprioritize data usage by heavy users in relation to other users’ data. Sandvine’s FairShare network management tool is intended to maintain the QoE for the average user while slowing, or “throttling,” the speed and increasing the latency of heavy users during periods of congestion much sooner than the average user would be affected (Sandvine, 2010). The issue of “throttling” download speeds down as a way to regulate behavior is one that has attracted a great deal of attention in the media portrayals of the network neutrality issue and by the FCC. In fact, the decision to discriminate on the basis of data packet content through speed regulation was one of the catalysts to the FCC’s attempted sanctioning of Comcast that led to the Comcast v. FCC (2010) Court of Appeals decision (Federal Communications Commission, 2008, p.1).
The potential ramifications for those who want to promote the causes of democracy in a society are varied. On the one hand, you have a provider discriminating on the basis of content, which has certain anti-social organizing potentials if ISPs begin discriminating on any ground they desire. Especially in a system in which there is little network transparency to determine the process that goes into speed regulation, providers could pick and choose to slow the arrival of data packets from certain sources that do not align with their own ideological perspectives. On the other hand, if ISPs increasingly begin inspecting packet contents, that could put them in the position to spy on citizens’ proclivities on behalf of the government, which certainly would be an anti-democratic outcome. This type of move on their part illustrates many of Morozov’s (2011) concerns regarding the illegitimate government’s triangle toolkit of surveillance, censorship, and propaganda.
Cumulative Data Transfer Arrangements
Flat Rate Unlimited Usage
The current model for many wireline connections, including DSL and cable Internet, is one where users pay a flat fee for unlimited broadband usage. This approach is certainly appealing for many Internet users in terms of avoiding bill-shock since the cost for Internet access is the same every month. Users do not have to be concerned over whether they have streamed too many movies (a total data transfer of approximately four gigabytes (GB) for a two-hour high-definition movie) in conjunction with all other uses of their Internet connection, including Skype, gaming, etc.
However, while providing for a flat rate, unlimited usage model of Internet service provision would involve the least ISP-infringement and cumbersome regulations on behavior, data-hogs who use substantially more bandwidth than any other end-users have the potential to crowd out other, more moderate users. As such, this policy alternative does not have a very strong endorsement for democratic potential based on policy substance and framing, though it does not have a negative, inhibiting impact on democratic organizing.
Tiered Data Usage Packages
Tiered data usage plans have already been rolled out by many wireless data providers, encouraged by the substantial network congestion experienced by AT&T following the widespread adoption of Apple’s iPhone. For wireless broadband, a move away from unlimited usage is intended to encourage moderation. In the case of AT&T, the wireless service provider analyzed the usage patterns of its users and devised two data plan tiers that reflect the average usage of 68% of users (200MB/month) and 98% of users (2GB/month), respectively (AT&T website). Tiered data usage plans for wireline connections would likely have higher thresholds. As an example of how these packages for wire line connections would be rated, in 2009, Time Warner Cable proposed but did not implement a tiered cable Internet system with monthly data usage caps of 5GB, 10 GB, 20GB, and 40GB with overage charges of $1/GB (Lowry, 2009). Time Warner Cable discovered that even announcing tests of implementing a tiered pricing system is likely to generate outrage from both its consumers and the general public. They ultimately canceled the plan before ever reaching the implementation stage (Ewing, 2009).
A movement away from flat fee Internet provision has three main benefits. The first, as mentioned above, is that it encourages moderation by users. Instead of an all-you-can-eat buffet of data usage, end users would prioritize what data they want the most. This data could potentially be mobilizing over a cause or gathering information about an issue for deliberation, but there is no way to really know if the majority of users would chose to use their limited data allowance for the month on non-entertainment or non-commerce activities. However, user-prioritization based on a perception of scarcity could still greatly reduce network congestion, resulting in a better QoE for all users. The second benefit from a tiered data usage package is that users would actually pay just for their level of data usage, resulting in lower prices for many, because light users would no longer subsidize heavy users. Sandvine (Fall 2010) notes that while the median amount of data used in a month was approximately 4 GB/user in September 2010, the mean monthly data usage was just shy of 15GB/user. In its research, it was not uncommon to find users that regularly transferred more than 4 terabytes (TB) of data in a month. This discrepancy means that light users who mainly utilize the Internet for non-bandwidth-intensive services such as email, social networking, and VoIP are paying considerably greater prices per byte to support the provider’s infrastructure than heavy Internet users who may be accessing the Internet from multiple devices simultaneously to stream movies, download TV shows and music, and engage in peer to peer file transferring (Sandvine, Fall 2010). Lower Internet access bills could go a long way towards bringing in those individuals who are still left out and on the wrong side of the digital divide because of financial limitations. This would increase the number of individuals who have online deliberative participation available as a legitimate option, even if they choose not to embrace it. The third benefit from moving to a tiered access system is that service providers would be better able to generate revenue to reinvest in existing infrastructure to reflect the actual usage patterns of their customers.
This model does have a fair number of drawbacks to it, however. People have grown used to a model of Internet provision that is similar to cable television in that it does not matter how much or how little one uses it, one pays the same fee. Another drawback to this model is that as new bandwidth-intensive applications and services roll out, they are likely to require higher and higher levels of bandwidth and providers would need to prepare for the ability to adjust tiers as usage changes. Unfortunately for service providers, a dynamic pricing model to reflect such shifts is likely to result in intense scrutiny and possible sanctions by the FCC or other governmental agencies. Additionally, being the first to make the move to a tiered system is also likely to result in less-than-satisfied customers responding poorly to being hit with overage charges.
Pay-per-byte
Some participants in the debate argue that rather than setting pricing guidelines that lump large numbers of users together in such a way that discourages data usage in moderation, the fairer model would be to charge users per byte, similar to how electricity users are charged per kilowatt-hour. This would allow users to evaluate exactly how much a song download or a movie streamed is worth to them, which may actually have the effect of some individuals transitioning back to cable and satellite television if they are large consumers of entertainment as it would then be cheaper than to pay per byte. The price per byte would need to be revised as the cost of the technology goes down and infrastructure is expanded to better handle increased loads.
Cap plus Overage Charges
Many companies, including Comcast, Cox, and Verizon have implemented monthly data usage caps. For example, Comcast allows residential users to have up to 250GB of data transfers per month without penalty while Cox and Verizon set their limits lower. This is an attempt to regulate the heaviest users from unduly impacting the network to the detriment of other users. The motivations behind implementing overage charges for usage above the cap would be to ensure both that the heaviest users are not being subsidized by average users and that if heavy users consistently abuse the network, the company can generate additional funds to reinvest in the infrastructure to bolster network capacity. Providing funds through overage charges to increase capacity also helps to alleviate the concerns raised by the flat fee system of heavy users crowding out the average user by hogging bandwidth through file transfers and other resource intensive actions.
Proposed Hybrid Solution
In a perfect world, there would be unlimited quantities of bandwidth available so that all packets would be treated the same and there would be no impact on the end user’s experience. However, that is not a reality given the current infrastructure, so decisions must be made regarding some degree of prioritization of content. My goal in this article is to fashion a solution that would improve the overall QoE on the part of end users while maintaining the Internet’s overall freedoms. Additionally, the desires of the end user can be hampered by service providers not fully investing in improving infrastructure as demands for bandwidth-intensive applications increase. Some of the lack of capital investment by service providers is related to concerns with profits, which is only logical considering that for-profit companies provide the services in almost all cases. If we were able to go back in time and change how the Internet has evolved and who controls its physical infrastructure, Internet service connections might be provided by not-for-profit entities that recognize the ability of the Internet unfettered by economic interests and to open up the possibilities of participatory democracy and enable a free flow of information in society. However, my proposed hybridization of data usage management options is intended to work within the limits of the current system and not create an overly intrusive regulatory mechanism to control the actions of corporations, a key concern when attempting to garner further support from both sides of the debate.
Earlier in this article, I elaborated on some of the primary benefits and shortfalls of the proposed solutions. Taking those into consideration, I will strengthen the arguments for why the two proposals I selected will provide the greatest benefits to the most people: consumer-driven prioritization and a usage cap plus overage charges. The consumer-driven prioritization approach acknowledges that not all data packets are equally important to the overall QoE. In all three of the qualitative data management approaches discussed here, the quality of the end user’s experience is considered the most important in judging the overall success of a network management strategy (Sandvine, 2008; Center for Democracy and Technology, 2010; The Free State Foundation, July 2010; Cox Communications, 2010). Sandvine (Fall 2010) notes “network policies generally exist to protect the experience of the vast, vast majority of subscribers” (p.14). If a data packet during transmission of a Skype VoIP phone call does not arrive in a timely manner (defined as below 100ms) and in the correct order as the other packets, the result could be a choppy or disjointed voice that is hard to understand.
If two or more users are using the same end connection at the same time, they may have different needs. Consumer-driven prioritization allows those users to select among their own needs which of the competing time-sensitive applications they are using should have priority over the other. If VoIP is rated as a higher priority than movie streaming, those packets can be given priority for one user while the other user has a buffer zone built into the movie or song that allows for a certain degree of increased latency that is higher than VoIP and both have stronger needs for lower latency than a third user who may be engaging in peer-to-peer file transfers for later movie viewing. While this approach to prioritization may not actually reduce congestion rates, it will help to mitigate the effects on a user’s experience, which is actually more important. For many applications, a slightly lower connection speed, such as 10mbps instead of 15 mbps, makes less of a difference than a longer latency, such as the 500ms lag time of satellite Internet connections versus the 150ms lag time with cable Internet. This consumer-driven prioritization would be in effect at all times, but consumers would be able to adjust their preferences online at their discretion. Provider-directed “traffic shaping” would only be enacted for the worst periods of network congestion in which the load on the network surpasses its capabilities. Cox found in its pilot test of network management tools this past year that 59% of its customers experienced moderate congestion that potentially impacted their QoE, and 10% experienced congestion lasting 30 minutes or more a day (Cox Communications, 2010, p.29).
My recommendation regarding cumulative data transfer limits is that a floating threshold based on the cumulative data transfers per user of 90% of users from the previous month be set with a flat rate for all users falling under that threshold. The 10% of users that use a service the most in the month would pay for the additional load capacity incurred on the network, which is likely to be greater than or equal to the bottom 50+% of users. The top 10% of users would then need to evaluate how much their excess Internet consumption, which is likely to be file transfers, is worth to them. Overage charges could encourage them to exercise moderation of, or at least take into consideration, the number and types of files they download in order to prioritize what are truly important to them and worth having. Additionally, the extra charges can be used to reinvest in the infrastructure on a regular basis to ensure that congestion does not continue to appear due to older technology that cannot support the load demanded by users in an area. Lastly, charging overage fees to the top 10% of users ensures that those users closer to the median or even the mean data usage will not be subsidizing the Internet connections and data usage of the heaviest users.
These two approaches used in conjunction remove decision-making about packet importance from service providers and put the decision in the hands of the user. As such, consumer advocacy groups would be hard pressed to find cause to complain about ISP practices as long as they are open and transparent and make it easy for users to adjust their settings. Lastly, the threats to democracy that could be present in a scenario in which corporations or government make decisions of which kinds of content are valuable and which kinds are removable while still maintaining an ability for networks to be managed such that users will have the best possible experience, including if and when they gather information about companies, social movements, or legislation to better educate themselves on how their society operates.
Conclusion
Internet policy is a constantly evolving field in which another court decision or legislative move could either empower or remove authority from the FCC or another agency to regulate service provision. If the FCC is able to successfully articulate and frame a “third way” of conceptualizing the Internet, as they have proposed in some of their most recent rulemaking calls for comment, then these policy categories will be a useful starting point to seriously evaluate the complicated relationship of Internet technology, democracy, and society. However, even if Congress does not explicitly grant them authority and the FCC’s attempts to carve out the authority for themselves is ultimately shot down by another Appeals Court, the policies here can be promoted or disparaged by citizen’s groups and individuals who care about democratic potential in society can apply pressure to Internet service providers through both purchasing power and awareness campaigns. If interested parties, myself being one of them, can publicize that the issues being discussed are not as simple as black and white of total restrictions and total freedoms, then maybe we have a chance to use these technologies both as catalysts and tools for actual public deliberation, taking our society one step closer to Habermas’s vision of a deliberative democracy in which we start with an open agenda to consider issues without manipulation in a forum that is open and easily accessible to any who wish to participate (Froomkin, 2003, p.869). Achieving such a vision will not be an easy journey, but I think that it is one we cannot afford to not take.
References
AT&T. Data plans with Wi-Fi: DataPlus and DataPro plan. Last accessed December 15, 2010 from http://www.att.com/shop/wireless/plans/data-plans.jsp
Becker, S., M. D. Crandall, K.E. Fisher, B. Kinney, C. Landry, & A. Rocha. (March 2010). Opportunity for all: How the American public benefits from Internet access at U.S. Libraries. Institute of Museum and Library Services. http://impact.ischool.washington.edu/documents/OPP4ALL_FinalReport.pdf
Broache, A. (October 29, 2007). Obama pledges Net neutrality laws if elected President. CNet News. Last accessed April 20, 2011 from http://news.cnet.com/8301-10784_3-9806707-7.html
Center for Democracy and Technology. (April 26, 2010). Reply comments of the Center for Democracy and Technology in the matter of preserving the open Internet (GN docket no 09-191 and broadband industry practices (WC docket no 07-52). Last accessed December 11, 2010 from: http://www.cdt.org/files/pdfs/CDT_Reply_Comments-Open_Internet.pdf
Comcast Communications. XFINITY Internet from Comcast. Website last accessed on December 14, 2010 from: http://www.comcast.com/Corporate/Learn/HighSpeedInternet/highspeedInternet.html
Cox Communications, Inc. (January 14, 2010). Comments of Cox Communications, Inc. in the matter of preserving the open Internet (GN docket no. 09-191) and broadband industry practices (WC docket no. 07-52).
Dahlgren, P. (2009). Media and political engagement: Citizens, communication, and democracy. New York: Cambridge University Press.
Ewing, J.B. (April 17, 2009). Time Warner Cable to shelve new Internet pricing plan. Last accessed December 14, 2010 from http://www.news-record.com/content/2009/04/16/article/report_time_warner_to_stop_new_pricing_plan_in_ny
Froomkin, A.M. (January 2003). Habermas@Discourse.Net: Toward a critical theory of cyberspace. Harvard Law Review 116(3), 749-873.
Federal Communications Commission. (August 20, 2008). In the matters of formal complaint of Free Press and Public Knowledge against Comcast Corporation for secretly degrading peer-to-peer applications and Broadband industry practices petition of Free Press et al. For declaratory ruling that degrading an internet application violates the FCC’s Internet policy statement and does not meet an exception for “reasonable network management.” Last accessed December 1, 2010 from http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-08-183A1.pdf
Federal Communications Commission. (October 22, 2009). Notice of proposed rulemaking: In the matter of preserving the open Internet broadband practices. Last accessed November 15, 2010 from http://hraunfoss.fcc.gov/edocs_public/attachmatch/FCC-09-93A1.pdf
Federal Communications Commission. (March 2010). Connecting America: The National Broadband Plan. Last accessed April 20, 2011 from http://download.broadband.gov/plan/national-broadband-plan.pdf
Federal Communications Commission. (June 17, 2010). In the matter of framework for broadband Internet service (GN Docket No. 10-127). Last accessed April 15, 2011 from: http://www.fcc.gov/Daily_Releases/Daily_Business/2010/db0617/FCC-10-114A1.pdf
Goldsmith, J. & Wu, T. (2006). Who controls the Internet? Illusions of a borderless world. New York: Oxford University Press.
Lowry, T. (March 31, 2009). Time Warner Cable expands Internet usage pricing. BusinessWeek. Last accessed December 15, 2010 from http://www.businessweek.com/technology/content/mar2009/tc20090331_726397.htm
Margolis, M., & G. Moreno-Riaño. (2009). The prospect of Internet democracy. Burlington, VT: Ashgate.
Mark, R. (November 21, 2007). John McCain: A Republican tech record. Government and Federal IT News. Last accessed April 21, 2011 from http://www.eweek.com/c/a/Government-IT/John-McCain-A-Republican-Tech-Record/
McChesney, R. W. (2004). The problem of the media: US communication politics in the 21st century. New York: Monthly Review Press.
Morozov, E. (2011). The net delusion: The dark side of Internet freedom. New York: PublicAffairs.
National Telecommunications and Information Administration (NTIA). (July 1995). Falling through the net: A survey of the “have nots” in rural and urban America. Washington, DC: US Department of Commerce, Retrieved April 5, 2009 from: http://www.ntia.doc.gov/ntiohome/fallingthru.html.
—. (1998). Falling through the net II: New data on the digital divide. Washington, DC: US Department of Commerce, Retrieved April 5, 2009 from: http://www.ntia.doc.gov/ntiahome/net2/.
—. (July 1999). Falling through the net: Defining the digital divide. Washington, DC: US Department of Commerce, Retrieved April 5, 2009 from: http://www.ntia.doc.gov/ntiahome/fttn99/contents.html.
—. (October 2000). Falling through the net: Toward digital inclusion. Washington, DC: US Department of Commerce. Retrieved from: http://www.ntia.doc.gov/ntiahome/fttn00/contents00.html.
—. (February 2010). Digital nation: 21st century America’s progress towards universal broadband access. Washington, DC: US Department of Commerce. Retrieved from: http://www.ntia.doc.gov/reports/2010/NTIA_Internet_use_report_Feb2010.pdf
Sandvig, C. (2007). Network neutrality is the new common carriage. info 9 (2/3): 136-147.
Sandvine. (Fall 2008). Fall 2008 global Internet phenomena report. Last accessed November 5, 2010 from http://www.sandvine.com/news/global_broadband_trends.asp.
Sandvine. (2010). Sandvine FairShare traffic management. Last accessed December 13, 2010 from www.sandvine.com/…/Sandvine%20Fairshare%20Traffic%20Management.pdf
Sandvine. (Fall 2010). Fall 2010 global Internet phenomena report. Last accessed November 5, 2010 from http://www.sandvine.com/news/global_broadband_trends.asp.
Selwyn, L.L., and Golding, H.E. (December 2010). Revisiting the regulatory status of broadband Internet access: A policy framework for net neutrality and an open competitive Internet. Federal Communications Law Journal 63(1), 91-140.
Stevenson, S. (2009). Digital divide: A discursive move away from the real inequities. Information Society, 25(1), 1-22.
The Free State Foundation (July 15, 2010). Comments of the Free State Foundation: In the matter of framework for broadband Internet service (GN Docket No 10-127). Last accessed December 12, 2010 from: http://freestatefoundation.org/images/Framework_for_Broadband_Internet_Service_-_Initial_Comments_071510.pdf
United States Court of Appeals. (April 6, 2010). Comcast Corporation vs. Federal Communications Commission and United States (No. 08-1291). Last accessed April 19, 2011 from: http://www.cadc.uscourts.gov/Internet/opinions.nsf/EA10373FA9C20DEA85257807005BD63F/$file/08-1291-1238302.pdf
Valadez, J.R., & R. Duran. (2007). Redefining the digital divide: Beyond access to computers and the Internet. The High School Journal.
Verizon Communications & Google, Inc. (August 2010). Verizon-Google legislative framework proposal. Last accessed December 15, 2010 from http://www.google.com/googleblogs/pdfs/verizon_google_legislative_framework_proposal_081010.pdf