purple motes

Alone in bed, her black hair untied across her pillow, she laments the white snow piling up in it. The open-faced puppeteer moves with her, their gestures double, her loneliness is his loneliness.

In many forms of puppetry, the puppeteer isn't visible. In Bunraku, a famous Japanese style of puppetry, three puppeteers visible on stage operate one puppet. Kuruma ningyo, a less well-known Japanese style, has only one puppeteer operating the puppet. The puppeteer sits on a cart and uses his legs to move the puppets legs, his left hand to control the puppet's left hand, and his right hand to control the puppet's right hand and head. The visible puppeteers are clothed in black, and often they were black masks. So, even when the puppeteer is visible, he or she is effaced as a person.

Nishikawa Koryu V performed puppetry for the song Kurokami (Black Hair) using Hachioji kuruma ningyo at the Freer Gallery yesterday. His face was exposed as he controlled the puppet and moved with the puppet. The doubling of the puppet and the puppeteer was a beautiful aspect of the performance.

3 Comments

Homes in Japan are rapidly being connected to optical fiber communications networks. Japan's incumbent communications company, NTT, had 6.08 million subscribers to its fiber network at the end of March, 2007. That figure indicates an increase of 2.66 million subscribers from March, 2006.[1] NTT's share of fiber network subscribers in Sept., 2006, was 66%.[2] If that share held constant through March, 2007, total fiber network subscribers in Japan at the end of March, 2007, was 9.2 million.

NTT's plans suggest that all homes in Japan will be connected to fiber networks by 2010. NTT East already has fiber reaching at least 75% of homes in its service area.[3] NTT as a whole seeks to have 30 million fiber network subscribers by 2010.[4] Japan has about 50 million households in total. NTT's plans are thus consistent with, in 2010, it having a 60% share of fiber network subscribers and all households in Japan being connected to fiber networks.

Having ubiquitous fiber networks in Japan in 2010 is consistent with the goals the Japanese Telecommunications Council set in a 1994 report. That report declared a target date of 2010 for "nationwide upgrading of fiber-optic networks":

in response to growing demand, the full-scale development and final expansion of network infrastructure should be promoted toward accomplishment by the year 2010. Fiber-optic cables should be installed at least on feeder lines, up to the distribution points, to promote the creation of a situation where infrastructure can be readily utilized at each stage.[5]

The associated network diagram shows optical network units attached to each house and business.

NTT has invested heavily in its fiber network while being subject to comprehensive unbundling regulations. NTT is required to unbundle copper loops and to allow line sharing (competitive DSL service over the primary copper loop NTT uses to provide plain old telephone service (POTS)). NTT is also required to unbundle fiber loops and interoffice fiber.[6]

Prices for unbundled elements seem to favor copper unbundling much more than fiber unbundling. The per month prices for sharing an unbundled cooper loop and for an unbundled fiber loop are about JPY120 and JPY5,100, respectively. At USD1=JPY117, the dollar equivalents are $1.03 and $43.59 per month, respectively.[7] NTT's market share for DSL subscribers is 39%, compared to 66% for fiber subscribers.[8] Given that the NTT unbundled copper loop charge is much lower than the NTT unbundled fiber loop charge, it makes sense for competitors such as Softbank (Yahoo! Broadband) to focus on providing DSL service over NTT copper. Moreover, given the low copper loop rates, it's not surprising that competitors use NTT copper loops rather than installing their own copper facilities.

NTT asserts that regulated unbundled fiber prices are much lower than NTT's costs of providing those facilities, but relevant data are difficult to interpret. Unbundling can imply a wide variety of charges in addition to monthly loop recurring charges. Ties between charges for using unbundled elements and other charges, such as universal service charges and POTS charges, can be highly relevant for evaluating unbundling economics. With respect to just unbundled loop costs, Takashi Ebhihara, a senior NTT executive, recently stated that NTT's costs for fiber loops were $200 per month, but with increased volume have fallen to $100 per month.[9] Because of economies of scope in laying fiber, costs per subscriber depend strongly on the share of eligible subscribers who actually subscribe. That may be part of the explanation for NTT's reported falling cost.

Even $100 per month costs for fiber loops is high. Verizon reported that its total capital expenditure in 2006 per home passed for its fiber network expansion was about $900, with about an additional $900 per home connected.[10] Fiber is an asset with a long life, and Japanese interest rates are nearly zero. At cost of $100 per month over just five years at zero interest, the present value of the cost is $6000. Moreover, NTT sells FTTH and FTTB at JPY6,700 and JPY3,950 per month (equiv. $57 and $34 per month), respectively.[11] These prices, which are much lower than the quoted loop costs, make sense only if NTT is aggressively seeking to acquire fiber subscribers and expects its fiber network costs to fall significantly in the future.

The two most important factors accounting for the expansion of fiber networks in Japan are probably very low monetary interest rates and institutional policy commitment to provide ubiquitous fiber network. Since the mid-1990s, the Japanese central bank discount rate has been below 1%. In contrast, U.S. Federal Reserve discount rate has been above 5% for most of that period. Cheap access to capital encourages major capital investments.

A ubiquitous fiber network has been a broad policy goal for both the Japanese government and NTT. The 1994 Telecommunications Council Report clearly expresses the importance the Japanese government associated with communications network investments. With respect to unbundled element prices and costs, Ebihara observers, "It's a huge issue. It's very difficult." (watch above video from time 38:20 to 42:10) At the same time, Ebihara clearly believes that investment in fiber networks has general importance for all citizens and for NTT's future. In a poignant conclusion to his presentation, Ebihara stated that he thinks NTT should both continue to lobby the government for higher unbundled fiber prices and continue to invest in fiber (watch video 1:24:40 to 1:32).

Update: Ernst-Olav Ruhle of Juconomy clarified that the rate for an unbundled copper loop cited above is the rate for line-sharing. The current rates for full unbundling for NTT East and NTT West copper loops are JPY 1,311 and JPY 1,393 per month respectively. The current rates for line-sharing for NTT East and NTT West copper loops are JPY 93 and JPY 101, respectively. Note that DSL providers such as Softbank (Yahoo Broadband!) provide voice service via VoIP, e.g. BB Phone. Hence they have no reason to purchase a full unbundled loop other than to avoid technical coordination problems with NTT.

Notes:

[1] See "NTT broadband fiber optics outshine ADSL," Asahi Shimbun, 04/04/2007. In Korea at the end of Feb. 2007, the Korean incumbent, KT, had 1.69 million fiber optic subscribers, Hanaro, 930,000 thousand, and LG Powercomm, 590,000, giving a total of 3.3 million FTTH subscribers in Korea. KT recently announced plans to bring fiber to every Korean household by 2010. The U.S., in contrast, had an estimated 1.3 million FTTH subscribers in Mar. 2007. Verizon, a leading investor in fiber communications networks in the U.S., had only 687,000 U.S. FiOS (fiber Internet data service) customers at the end of 2006. See p. 16 of Verizon's 4th quarter 2006 earnings presentation. For a review of that data, see Nyquist Capital.

[2] Japanese Ministry of Internal Affairs and Communications (MIC), Dec. 2006, as described on p. 12 of Takashi Ebihara, "The Japanese Broadband Miracle," presentation at the Information Technology & Innovation Foundation, Apr. 4, 2007.

[3] Ebihara, p. 11, citing NTT East data as of Mar., 2006.

[4] See Asahi article.

[5] From unofficial and tentative translation of Reform toward the Intellectually Creative Society of the 21st Century, Report, May 31, 1994, Goals for the Installation of Network Infrastructure. I'm grateful to Frank A. Coluccio for pointing out this report.

[6] Ebihara, p. 21.

[7] See Ema Tanaka, Minoru Sugaya, Sayaka Shiotani, Evolution of IP Network and Convergencein Japan – Impact of Hard law and Soft Law, presentation at ITS Conference in Beijing, 2006.06.13, slides 23-4 of powerpoint presentation. Ebihara, about 38:20 to 41:00, states that unbundled fiber loops cost about $50.

[8] Ebihara, p. 12.

[9] Ebihara presentation video above; watch from time 38:20 to 42:10.

[10] See Verizon Communications' FiOS Briefing Session, Sept. 27, 2006, pp. 24-5.

[11] Ebihara, p. 10.

2 Comments

In an infoworld article entitled "Government policies add to Japan's broadband success," Grant Gross led with this news:

A wide-ranging government policy on broadband and healthy competition among providers gives Japanese customers greater speeds at a much cheaper price than U.S. customers pay, a Japanese telecom executive said Wednesday.

Japanese customers pay about US$0.70 for each megabit per second of bandwidth, compared to $4.90 per megabit on average in the U.S., said Takashi Ebihara, senior director of the corporate strategy department at NTT East Corp. and a visiting fellow at the Center for Strategic and International Studies, a Washington, D.C., think tank.

In an article in Silicon Valley Watcher, Richard Koman had a similar article headlined, "Why broadband is 5x cheaper in Japan." Government policy is undoubtedly extremely important for the communications industry (and government bureaucrats deserve much more credit than they usually receive). A comparison between U.S. and Japanese bandwidth prices can be easily used to support conventional views about government policy and competition. Such a comparison can also be ignored if more convenient for a particular point of view.

Ebihara's presentation was much more interesting than these news reports indicate. Ebihara actually compared bandwidth prices across twelve countries. Moreover, he cited a source for the data quoted above. His source was ITU Internet Reports 2005.[1] The table below includes all the relevant data given in that source. Thinking about policy and competition with respect to this set of countries and the range of prices that exists suggests that bandwidth prices are not strongly correlated with objective industry structures.

NTT's current broadband service prices do not have a consistent bandwidth price level. ADSL has a price per megabit about three times higher than the price per megabit for Fiber To The Building (FTTB -- used for multi-tenant buildings). Fiber to the Home (FTTH) is about two-thirds more expensive than FTTB. NTT's ISDN, on a per megabit basis, is about two thousand times more expensive than its FTTB.

Bandwidth is more meaningful as a technical characteristic of a widely available service than as a good that users individually purchase. Most communications service users have little understanding of the concept of bandwidth. Most communication service providers do not guarantee the bandwidth of services purchased, nor define clearly what the nominal bandwidth of the service means. Moreover, the bandwidth of a "connection to the Internet" is no more meaningful than the bandwidth of a "connection to connections".

A communications service business can be insightfully divided into two important activities. Building more capable communications networks and migrating users to them is one important activity for a communications business. In Japan, that is what NTT has done in shifting subscribers from ISDN to ADSL and then to fiber. Acquiring funds is another important activity for a communications business. Prices per megabit do not provide a good connection between these two aspects of a communications business.

OPLANs provides a useful alternative perspective on bandwidth prices. OPLANs emphasizes charging for access, not bandwidth. Thus an OPLAN is meant to be:

a network of truly ‘broadband’ capacity - i.e. where the bandwidth capacity is dictated by nothing other than physical characteristics of the deployed technologies [2]

With an OPLAN, users get to use as much bandwidth as they can. With modern fiber optics, that's very high speed without any price. Moreover, that doesn't depend on any particular national government policy, nor depend on competition.

Notes:

[1] See Taka Ebihara, Understanding the Japanese Broadband Miracle, p. 6, citing ITU Internet Reports 2005, p. 15.

[2] Malcolm Matson, "So What is an OPLAN?"

Offer Comments

Good, large-sample data on Internet traffic is hard to find. However, an excellent study of Internet traffic in Japan [Cho et. al. (2006)] describes 42% of Japanese (public) Internet traffic for six one-month observations spanning Sept. 2004 to May 2006.

This study reports many important facts. The ratio of bytes sent to residential customers to bytes received from residential customers was 1.3 -- remarkably symmetric. Peer-to-peer applications using dynamic port assignment account for most of the traffic. The distribution of bandwidth use by users is a high-powered power law. Most importantly, specific users' positions in that distribution vary over time and are not well identified with demographic and other customer characteristics (other than fiber connection, i.e. bandwidth availability).

The residential traffic growth figures for Japan vary considerably across month intervals. From Sept. 2004 to Nov. 2004, total traffic (aggregated inbound and outbound) grew 180% on an annualized basis. In subsequent 6-month intervals, annualized traffic growth rates were 58%, 19%, and 37%. Data on growth in fiber-to-the-home (FTTH) connections and total connections also shows volatility, but this does not seem to explain the reported traffic growth volatility. Any ideas about what explains it?

According to my calculations, (public) Internet traffic in Japan from Sept. 2004 to May 2006 grew about 60% per year. This figure aggregates inbound and outbound traffic and includes residential broadband customers and non-residential broadband customers (leased lines, data centers, dialup). I estimated non-residential traffic for the 7-ISP sample using the reported residential/non-residential traffic figures from the 4-ISP sample in Nov. 2005, and the non-residential broadband growth rate for the 4-ISP sample from Sept. 2004 to May 2006.

This estimated annualized growth of Internet traffic in Japan is much higher than estimated annualized growth of (non-voice channel) bandwidth in use in the U.S. across the 1990s. From 1989 to 1999, total DDS, DS1, and DS3 channel termination bandwidth in the U.S. grew an estimated 27% per year (see Table P6 in U.S. Bandwidth Price Trends in the 1990s). The difference between a 60% growth rate and a 27% growth rate year becomes huge after only a small number of years.

The Japanese study does not encompass bandwidth deployed in private networks. The ratio of Internet bandwidth to total inter-office bandwidth may have been about 15% in the U.S. in 1998 (see Growth in the "New Economy", p. 6). The Japanese data show much a faster growth rate of non-residential bandwidth than residential bandwith, but the former is only about two-thirds the size of the latter. This is consistent with a large share of non-residential bandwidth not being incorporated into the public Internet.

An important point: the re-organization of network transmission protocols on existing networks tends to occur very slowly. Astonishing fact: in the U.S., about 90% of mobile-phone communications towers use traditional, copper-based TDM backhaul. Legacy networks hang around for a long time.

Reference:

Cho, Kenjiro, Kensuke Fukuda, Hiroshi Esaki, and Akira Kato, "The Impact and Implications of the Growth of Residential User-to-User Traffic," sigcomm 2006.

Offer Comments