Tuesday, May 24, 2011
Refuting the FCC’s Self-Serving Optimistic Conclusions about Reverse Auctions with International Examples of Challenges and Failures
Since I started Rural TeleCommentary in January of this year, I have continually expressed my frustration with the FCC's proposal to use reverse auctions (also known as "incentive auctions") to distribute Universal Service Funds. I first discussed my opposition to reverse auctions in a post about the Mobility Fund, and I have consistently covered them throughout my series on USF Reform. Today, I am going to discuss one of the main reasons why the FCC should be wary of implementing reverse auctions on a large scale: reverse auctions have never been used in the U.S. telecommunications industry, and have failed spectacularly in other countries. I mentioned last month that I wanted to analyze some international telecommunications policies, and I believe this is one case where the FCC really needs to pay serious attention to how reverse auctions have been used—and how they have failed—in other countries, like Chile and India. Since there is no local data or models for telecom reverse auctions in the U.S., there is actually no choice but to look beyond our borders for information. Based on the following studies, I feel strongly that the international telecommunications industry is sending very clear signals that reverse auctions will fail to achieve the goal of distributing universal service funds efficiently and to the most capable provider, but this is just my interpretation of the limited information available (which is predominately supportive of reverse auctions).
Irene Wu from the FCC's International Bureau seems to interpret things differently, but one has to wonder how self-serving her October 2010 report, "Maximum Impact for Minimum Subsidy: Reverse Auctions for Universal Access in Chile and India," is. She concludes that reverse auctions were successful in these countries and will be a fast and efficient way to distribute USF here in the U.S. When I first read this paper last year, I did not see very many indications of "success" with reverse auctions in India and Chile, and it is not unusual for the FCC to release self-serving data prior to making a major decision that the data supports. Case in point, the Seventh Broadband Progress Report and Order on Reconsideration (706 Report), was released at the end of the day last Friday, and has already received considerable criticism for being self-serving and an excuse to impose heavy-handed regulations to ensure broadband is deployed at a more rapid pace going forward. The Report concluded that 100 million Americans do not have broadband, with 26 million Americans having no access to broadband. Here is the shady part—they use data from the NTIA's National Broadband Map which has been widely disputed and deemed inaccurate. This is not to say that there aren't millions of Americans without broadband and we need a USF mechanism to build out networks quickly, efficiently and relatively inexpensively, but the data used in this study may not tell the whole story. The report concludes that broadband is not being deployed fast enough, but I personally don't think "fast enough" is a reliable term of measurement. Fast enough compared to what? Anyway, according to Hillicon Valley, "The report is controversial because industry sees it as a pretext for regulation. The FCC used last year's negative findings to help justify net neutrality regulations passed in December. Rep. Greg Walden (R-Ore), chairman of the House telecom subpanel, made it clear Monday that he thinks the findings are without merit. 'It is one thing to recognize that some areas of the country—typically rural—are difficult to serve; it is quite another to say that broadband is not being reasonably and timely deployed to all Americans,' he said" (Jerome, Hillicon Valley, May 23, 2011). Not to stray off topic too much, but this report, and Wu's report on reverse auctions, are both examples of the FCC releasing self-serving data to justify controversial regulatory decisions.
Wu's report identifies requirements for successful reverse auctions, such as clear policy targets, methods for calculating how much money is available for each project, minimum technical and financial requirements, procedures for distributing funds, increasing participation, and evaluating bids (Wu, FCC, 2010, pg. ii). However, the FCC and participating members of the FCC's Workshops on USF Reform have not been able to come up with any specific or attractive proposals to meet any of these requirements, literally. They cannot even decide on how large or small the auction blocks should be, or if providers of different technologies can participate in the same auction. I personally think it could take years for all these details to be worked out, which could make broadband deployment even slower, and further frustrating the FCC, who already thinks broadband is being deployed too slowly.
Chile has the longest history with reverse auctions, which started in 1995 as a method to deploy public pay telephones in rural areas, and they are still being used to deploy fiber and mobile infrastructure to unserved areas. Chile's telecom regulator, Subtel, identifies project areas by soliciting requests from communities, telecom providers and the public (Wu, pg. 4). In the early years of reverse auctions for payphones, Chile had a difficult time developing a specific, targeted definition for projects that should receive subsidies. This is something that the FCC is already struggling with, as they have not been able to define the size of areas that will be auctioned. Furthermore, Chile struggled with a version of the "donut and hole" issue that the FCC is facing—in Chile, low income "holes" within higher income "donuts" were ineligible to receive subsidies, leaving areas with the greatest need for telecommunications services behind in development. Wu cautions that, "to avoid these kinds of problems, substantial local data on population, geography, and other socioeconomic indicators should be collected in advance of making policy decisions" (Wu, pg. 5). This worries me because the FCC and NTIA cannot even get accurate data for the National Broadband Map, and they are planning to rely on this data to determine subsidies for unserved areas. In a major 2008 reverse auction to provide Internet to over 1,000 rural communities, another significant problem emerged: the winner's funding fell through and it could not build the networks. The winner, Inverca (a Chilean-Malaysian joint venture, until the Malaysian partner bailed), had to surrender its license to the runner-up, Entel. Although Entel ended up building out mobile broadband to the rural communities, it certainly delayed the desired outcome and serves as a major cautionary tale for the U.S. In this report, Wu sees the Chilean reverse auction process as being very successful, but I see it as being a long and bumpy road with questionable success.
India is no stranger to controversy in its telecom industry, and their reverse auctions were no exception. According to Wu, "years of stalled progress toward achieving universal service objectives spurred the Indian government toward implementing minimum subsidy auctions…By 2002, it was apparent that politically important goals, such as connecting a phone in every village, would not be achieved with the original plan to rely on competition in the market with universal service obligations attached to licenses" (Wu, pg. 11). Basically, India was warehousing USF and not using the funds to improve telecommunications service in rural and remote areas. The success of reverse auctions in India is highly dubious, and India's problems can definitely serve as a cautionary tale for the U.S. as well. India's incumbent wireline provider, BSNL, has made a clean sweep of most of the subsidies, just as many critics are anticipating will happen in the U.S. with companies like Verizon pushing small competitors and new entrants out of the running. In most of the auctions, BSNL bid the exact minimum benchmark subsidy, which eliminated viable competitors who may have been able to build the networks more effectively but at a slightly higher cost. In some cases, participants even submitted zero or negative bids, which is clearly not a representation of true costs but more likely a callous strategy to exclude competitors from gaining market share. After winning the most subsidies in a 2007 mobile infrastructure reverse auction, BSNL actually had the slowest rate tower installation using the subsidies. This is also a valid concern for the U.S., if a company like Verizon (or worse, AT&T) wins auctions that include rural geography surrounding a more densely populated community. How will the FCC enforce timely build-out in the rural areas, or ensure the winner is not just focusing on improving infrastructure in the most profitable areas of an auction block? This point has been raised repeatedly in the USF Reform proceeding comments and workshops, so it would be wise for the FCC to analyze BSNL's failures to use USF subsidies effectively and efficiently in the Indian reverse auctions.
Wu's paper is not the only source on international use of reverse auctions for telecommunications, thankfully. I also recommend "Reverse Auctions and Universal Telecommunications Service: Lessons from Global Experience," by Scott Wallsten of the Technology Policy Institute (2008, Federal Communications Law Journal), which is definitely less self-serving and includes analysis of a wider range of countries (such as Australia, Colombia, Nepal and Peru). Wallsten's paper is supportive of reverse auctions for universal service "if the regulator's goal is to reduce the level of subsidies or to provide information about the 'right' level of subsidies," but it also highlights the challenges that other countries have faced (Wallsten, pg. 4). Wallsten concludes that "global experiences reveal that auctions are feasible and that subsidies required are generally less than incumbents had previously led policymakers to believe," but I still do not find the examples in this report to be convincing reasons why the U.S. should implement reverse auctions (Wallsten, pg. 9). In a 2000 Australia reverse auction pilot program, none of the incumbent's (Telstra) competitors bid for the $85 million subsidies because they felt the subsidy was too low and they would not be able to compete with Telstra. This is basically my biggest fear for the U.S.—that small and regional providers will not even bother participating in reverse auctions, avoiding them completely as they often do in regular wireless license auctions. The dominant carrier simply intimidates them and the administrative expenses of participating may even be too high. Australia's goal in this auction was to increase competition, which it completely failed to achieve, possibly because subsides were insufficient to incentivize competitors to participate. Wallsten also addresses some of the challenges in Chile that were not discussed in Wu's paper, particularly the gaming done by the dominant provider: "The dominant local firm bid 100% of the maximum subsidy in areas with no competitors which were close to its existing network, 90% of the maximum subsidy in areas with an emerging competitor which were close to its network, and zero in areas with strong competition" (Wallsten, pg. 12). This is a really scary likely outcome of reverse auctions, and the FCC will need to consider mechanisms to prevent gaming or we may see a repeat of this ugly situation in the U.S. I personally do not think a provider should be allowed to place zero or negative bids; this action clearly signals that the provider can afford to build out the network without the subsidies, and the provider is mainly looking for the FCC to cement its status as a dominant provider. Zero and negative bids will completely undermine the goals of using reverse auctions to distribute USF subsidies in high-cost areas, as the subsidies would go to providers who clearly do not need the money in the first place. If a provider has the financial strength to bid zero or a negative amount, they need to just start building the network and not even bother participating in the auctions. Finally, Wallsten talks about reverse auctions in Peru, which occurred from 1999 to 2001 for providing telephone service in unprofitable rural areas. I thought this example was interesting because the number of telephones in Peru increased, but it is highly unlikely that this success can be attributed to reverse auctions but rather from greater global market forces resulting from liberalization and increased investments in telecom worldwide. Wallsten simplifies (possibly over-simplifies) the definition of reverse auctions by equating them to Request-for-Proposals (RFPs), which are commonly used by the government for both basic and complex products and services. Although Wallsten believes that reverse auctions can be successful, he offers some important words of wisdom: "One lesson is clear: details of the auction matter. A poorly designed auction may not generate any improvement over the status quo," (Wallsten, pg. 21).
I am worried that the FCC will hastily throw together the guidelines and requirements for reverse auctions because they are in such a rush to reform the USF system and make sure broadband is deployed more rapidly—the 706 Report illustrates the panic that the FCC is trying to induce regarding how the U.S. measures the success and failure of broadband deployment and adoption. There is simply not enough information available for the FCC to make informed and rational decisions when forming the rules for reverse auctions without first taking very small steps through extremely narrow pilot programs. The FCC should not approve reverse auctions in the USF Reform rulemaking because the risks are far too great.
I chose to revisit the Wu and Wallsten studies today because I am strongly considering an extensive focus on reverse auctions in my Master's thesis, so I wanted to write down some of my thoughts on how reverse auctions have failed in other countries. Readers, if you have any other interesting studies on reverse auctions in other countries, I would love to read them!
Tomorrow I will start reviewing the USF Reform reply comments, and I will hopefully have something published by Thursday. Also, I am finally making an effort to use LinkedIn, so please add me as a connection! I've had a LinkedIn account for years and have never really figured out how to use it effectively. I've had a lot of success so far with Twitter, so I figured I should give LinkedIn another shot.