Content uploaded by Masatsugu Tsuji
Author content
All content in this area was uploaded by Masatsugu Tsuji on Aug 19, 2014
Content may be subject to copyright.
econstor
www.econstor.eu
Der Open-Access-Publikationsserver der ZBW – Leibniz-Informationszentrum Wirtschaft
The Open Access Publication Server of the ZBW – Leibniz Information Centre for Economics
Nutzungsbedingungen:
Die ZBW räumt Ihnen als Nutzerin/Nutzer das unentgeltliche,
räumlich unbeschränkte und zeitlich auf die Dauer des Schutzrechts
beschränkte einfache Recht ein, das ausgewählte Werk im Rahmen
der unter
→ http://www.econstor.eu/dspace/Nutzungsbedingungen
nachzulesenden vollständigen Nutzungsbedingungen zu
vervielfältigen, mit denen die Nutzerin/der Nutzer sich durch die
erste Nutzung einverstanden erklärt.
Terms of use:
The ZBW grants you, the user, the non-exclusive right to use
the selected work free of charge, territorially unrestricted and
within the time limit of the term of the property rights according
to the terms specified at
→ http://www.econstor.eu/dspace/Nutzungsbedingungen
By the first use of the selected work the user agrees and
declares to comply with these terms of use.
zbw
Leibniz-Informationszentrum Wirtschaft
Leibniz Information Centre for Economics
Akematsu, Yuji; Shinohara, Sobee; Tsuji, Masatsugu
Conference Paper
Factors of FTTH deployment in Japan: A panel data
analysis
23rd European Regional Conference of the International Telecommunication Society,
Vienna, Austria, 1-4 July 2012
Provided in Cooperation with:
International Telecommunications Society (ITS)
Suggested Citation: Akematsu, Yuji; Shinohara, Sobee; Tsuji, Masatsugu (2012) : Factors of
FTTH deployment in Japan: A panel data analysis, 23rd European Regional Conference of the
International Telecommunication Society, Vienna, Austria, 1-4 July 2012
This Version is available at:
http://hdl.handle.net/10419/60361
1
Factors of FTTH Deployment in Japan: A Panel Data Analysis
Yuji Akematsu
Graduate School of Economics, Osaka University
Sobee Shinohara
Graduate School of Applied Informatics, University of Hyogo
KDDI and KDDI Research
and
Masatsugu Tsuji
Graduate School of Applied Informatics, University of Hyogo
Abstract
In accordance with the development of broadband services, interests are focused on the
deployment of optical fiber and various reports and researches are published, including
“Next Generation Connectivity” by the Berkman Center, Harvard University, and
“Indicators of Broadband Coverage” by OECD. These analyze the international
comparison of broadband penetration or how U.S. and EU lag behind. Less analysis was,
however, spent how Japan constructed the most successful FTTH networks all over the
country. The objective of this paper is to identify factors promoting deployment of
Japanese FTTH in the competitive environments utilizing the panel data analysis and
the prefectural panel data of subscribers. Major deregulations including unbundling,
openness of facilities, connection charges and collocations were already implemented.
The problem occurred in this process, however, is that the share of NTT locals in the
FTTH market locals has been increasing and reached more than 70%. Moreover, the
NTT locals’ share in the total four broadband markets including FTTH, cable modem,
ADSL and FWA becomes also more than 50%. This paper attempts to explain this
reason. Recently, however, the situation has been changing, namely competition has
been more severe in the metropolitan and populated areas, which is promoted by power
companies and KDDI. This paper also focuses on current changes in the FTTH market,
and explains how these are different from the previous period. Lessons learned from
Japanese experiences will provide good reference to other economies to deploy FTTH.
Keywords: FTTH; broadband connectivity; panel data analysis; unbundling; migration;
HHI.
2
1. Introduction
FTTH (Fiber To The Home) has already covered residential areas by nearly 100%
in the main part of metropolitan areas in Japan, and no other countries all over the world
can be found that have such diffusion. FTTH can not only transmit large amount of data
instantly, but also provide voice and video services with one single line (Triple-play).
Therefore, it will be expected to contribute various industries other than
telecommunications. Heavy users of the Internet have already migrated from ADSL to
FTTH, and the former subscribers are declining. In addition, NTT East and West
announced to start new broadband services based on their NGN (Next Generation
Network) in April 2008, which make it possible to transmit high speed and large volume
of images with guaranteed quality (NGN is not of best effort). This is the one we
waiting for in the age of broadband, and NGN becomes possible in the existence of
FTTH.
1.1. Objectives of the paper
Japan has taken a worldwide lead in FTTH, with 17.8 million subscribers as of
March 2010. Japan has been the first country to experience this transformation. Based
on this development, NGN services started in 2008. This paper analyzes factors
promoting Japanese FTTH. Tsuji and Tomizuka [2006], Akematsu [2008a], [2008b],
and Tsuji and Akematsu [2010] summarized factors promoting Japanese broadband as
follows: (1) deregulations by the government; (2) competition among carriers; and (3)
technological development. The paper identifies exactly which factors facilitating
Japanese FTTH development, especially focuses on how subscribers migrate from
ADSL to FTTH. In so doing, this paper utilizes the panel data analysis, and due to the
availability of related data, instead of subscribers of carriers, the prefectural panel data
of subscribers are used. In estimation, explanatory variables which promote FTTH
development are the number of cable modem and ADSL subscribers as competing
broadband technology, and the number of FTTH carriers operating particular prefecture.
Since Tsuji and Akematsu [2010] analyzed FTTH diffusion using prefectural panel data
models (pooled data model), this paper separates the nation into two parts, the Eastern
and Western Japan and attempts to identify factors promoting Japanese FTTH. This
paper thus aims at explaining how Japan takes lead in FTTH services.
1.2. Literature review
With respect to studies on the broadband diffusion factors, there have been
3
various opinions and discussions on the government policies such as deregulations for
broadband diffusion and competition, business strategies of operators, and attributes of
each country, in the scope of one single country or region, or multiple countries.
For studies on broadband in a single country, one example can be found in the
analysis on broadband markets conducted by the US Federal Communications
Commission (FCC [2010]), which focuses on the level of education, income, and other
attributes found in multiple areas throughout the US. Similarly, Tsuji and Akematsu
[2010] attempted to identify factors promoting Japanese FTTH using panel data analysis
based on prefectural data by focusing how subscribers migrate from ADSL to FTTH.
Other studies on the Japanese ADSL market include Tsuji and Tomizuka [2006] and
Akematsu [2008a] and [2008b] made positive analyses of ADSL diffusion factors,
suggesting that the driving force of ADSL diffusion throughout Japan is the open-access
policy for metal subscriber lines.
For studies on broadband, within a region as shown by FCC [2010] and
Commission of the European Communities [2009], or among several countries, as
shown by Kushida [2006] and [2009], Fiona [2009], and Scott [2009], one of the most
prominent works is the research conducted by the Berkman Center for Internet and
Society, Harvard University (Berkman Center [2010]). The research analyzes a wide
range of broadband diffusion factors, including competition-related issues such as
diffusion rates, access speeds, costs, etc. as well as government policies for broadband
diffusion and competition, operators’ investments, attributes of the country and other
factors. The Center examines a broadband diffusion policy including state aid in each
country, while studying broadband competition policies and suggesting a correlation
between countries with the open-access policy (unbundling, etc.) for subscriber lines
that are irreproducible and high broadband diffusion rates.
This paper consists of four sections: Section 2 describes the FTTH market in
general. In Section 3, data for estimation is summarized and also this section formulates
our methodology of estimation and presents the results. Discussions and brief
conclusions including some prospects of this research are provided in Section 4 and 5.
2. FTTH MARKET
Here the current situation of Japanese FTTH market is present as a premise to the
analysis.
4
2.1. Broadband market in Japan
One of the defining characteristics of broadband services in Japan is its sequential
development, which began with PSTN (analog telephone network) and ISDN before the
launch of broadband services, then moved to broadband systems such as cable modem
and DSL, and later on to FTTH, as Figure 1 shows.
The trend of subscribers of broadband is summarized in Figure1. The first
transformation of broadband was right after year 2000, when ADSL services was
introduced. Due to the deregulations such as unbundling, collocations, market
competition promoted sharply an increase in ADSL. In the middle of 2000s, the number
of subscribers of FTTH was increasing rapidly, while that of ASL showed peak out and
started declining, which was termed by “migration from ADSL to FTTH.” The number
of FTTH subscribers exceeded those of ADSL at the end of June 2008, that is, FTTH
comes to be major in Japanese broadband, which was the second transformation of
broadband. The number of subscribers of CATV is slightly increasing.
Figure 1: The number of broadband subscribers in Japan
Source: Ministry of Internal Affairs and Communications, and various company publications
(Unit: 1,000)
0
5,000
10,000
15,000
20,000
25,000
2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011
DSL FTTx CATV
(
BB
)
(Unit: 1,000)
20,300
8,250
5,680
5
2. 2. FTTH market
Different from the ADSL market, one of the issues related to Japanese FTTH is
dominance of NTT locals in the market. There are two sub-markets of FTTH;
condominiums and single house. Among approximate 17.8 million subscribers of the
whole market, the former amounts to 10.1 million subscribers (59.4%), while the latter
to 7.2 million (40.6%). As for the FTTH market as a whole, the share is as follows: NTT
locals (74.42%), power companies (9.2%), KDDI (8.0%), and USEN (3.0%) as of
March 2010. The market share of condominiums is NTT locals (70.1%), KDDI (8.0%),
USEN (6.7%), power companies (5.8%), while in the market of single house, NTT
locals (77.56%), Power companies (11.5%) and KDDI (8.0%) as of March 2010. NTT’s
share has been gradually increasing, and it is expected that NTT dominates the market.
A big difference regarding competition in the FTTH market in comparison with
ADSL is that carriers have to deploy optical fiber to single houses or condominiums by
themselves, since there can be no line sharing which is popular in the ADSL market. In
2005, the “shared access system” was introduced, in which carriers could borrow NTT
locals’ optical fiber network. This system, however, does not work as well as the line
sharing of ADSL. Thus, more deregulation is required to promote competition in the
FTTH market. The Japanese success of ADSL development is due largely in part to a
good cycle of deregulation and effective market competition: deregulation caused
competition, while competition demanded further deregulation. After all, subscribers
(consumers) can obtain economic benefits. This is a lesson well learned from the
Japanese ADSL experience, but one not yet fully applicable to this market.
Deregulation was almost completed in the early stage of ADSL. Openness of
copper lines was also applied to FTTH subscriber lines, and Japan is an only country to
admit this, while other counties do not since it harms for carriers to invest in FTTH
subscriber lines. The owners of FTTH networks such as telecommunication carriers,
power companies, CATV, water subways, etc. are obliged to open their networks to
other competitors. The openness is aimed at introducing service-based competition in
the FTTH market. The competitors’ use of NTT’s FTTH subscriber line remains only
0.5 percent, which is not similar to ADSL case. Power companies and KDDI, for
example, entered the market by constructing their own FTTH subscriber networks, and
thus even in Japan, service-based competition is not fully promoted.
The Implementation and events of the FTTH market which contributed to
diffusion is summarized in Table 1. Most of implementations and events are related not
only FTTH, but also ADSL, or other broadband technology. Therefore, many
6
deregulations have implemented before FTTH services started, and thus it may be
difficult to extract factors related to deregulation which promote FTTH.
Table 1 Implementation and Events of FTTH Market
Time Events
Revision and enforcement of Telecommunications Business Law
Enforcement Rule and connection fee rule
Apr. 2001
(Unbundling of dark fiber)
Enactment of notification
Jun. 2001
(Stipulation of information on collocation be disclosed free of charge)
Inauguration of the Telecommunications Business Dispute Settlement
Commission
Nov. 2001
(Mediation and arbitration became possible)
Sep. 2005 NTT started optical IP telephony service (“Hikari-Denwa”).
Mar. 2008 Revision of connection fee rule of dark fiber.
Source: MIC and NTT
The factors which effect FTTH diffusion are represented by migration,
competition among broadband technology, and charges. The other factors and the data
for estimation will be summarized in the next section.
3. ESTIMATION
In the previous paper, Tsuji and Akematsu [2010] extensively analyzed the FTTH
market such as market as a whole, single house and condominium sub-markets using
prefectural data by focusing the migration from ADSL to FTTH. This paper, on the
other hand, attempts to examine the difference between the Eastern and Western Japan
in the same framework.
3.1. Model specification and data
As mentioned, this paper attempts to estimate factors promoting FTTH in two
geographical markets, the Eastern and Western Japan. The motivation of the analysis is
that FTTH market is quite different in two part of Japan. In Eastern Japan, there is no
strong competitor, and accordingly the share of NTT East has been increasing, while in
Western Japan, power companies have been challenging NTT West which share is
relatively lower than NTT East. Thus, we hypothesize the latter is more competitive
than the latter. By estimating two models separately, this paper clarifies the factors
promoting Japanese FTTH.
7
In addition, the FTTH market can be divided into two sub-markets such as single
house and condominium and in terms of competition these two markets have different
characteristics. These two markets are examined separately. As already mentioned, NTT
locals have the majority of market share, not only whole FTTH market but also two
sub-markets. This paper also examines separately subscribers of NTT locals.
Accordingly, we estimate the following estimations: (i) Eastern Japan (whole, single
house and condominium markets); (ii) Western Japan (whole, single house and
condominium markets); (iii) NTT East (whole market); and (iv) NTT West (whole
market).
The method we use here is panel data models using prefectural data, after
estimating fixed effect model and random effect model, the best model is selected
according the F test, the Lagrange Multiplier test, and the Hausman test. The genera
model for estimation in common for these markets is formulated as follows.
)),0(~(
)ln()ln(
)ln()ln()ln()ln()ln(
2
65
43210
vititiit
it
Inter
it
Intra
it
itittitit
iidvvu
uHHIHHI
ADSLCATVGDPpriceFTTH
where FTTH
it
, ADSL
it
, and CATV
it
denotes the number of subscribers of each broadband
technology (per 100 people), price
it
monthly subscription charges (per 1Mbps and
normalized by Consumer Price Index of each prefecture), GDP
t
real GDP (also
normalized by CPI), Intra-HHI
it
and Inter-HHI Herfindahl-Hirschman Index of the
FTTH and broadband market, respectively. An error term is assumed to consist of
one-way fixed effect where λ
t
is a fixed effect and v
it
~ iid (0, σ
v
2
). Finally, i denotes
prefecture and t quarter. Related to HHIs, there are only 4 biennial data for each region,
and these data are transformed to anural then to quarter data by approximating them
linearly. The sample period is from the third quarter of 2005 to the last quarter of 2011.
The explanatory variables are the number of cable modem subscribers as
alternative technology, and the sign of cable modem coefficient is assumed to be
negative. Intra-HHI and Inter-HHI present the degree of competition in the market. The
more carriers enter the market, the more subscribers grow due to intense market
competition. But one carrier dominates the market, and the number of subscriber would
not increase. The signs of HHIs are expected to be negative. On the variable of ADSL,
8
the more subscribers one region has, the more people migrate to FTTH in that region
naturally.
The sources of data and summary statistics are summarized in Table 2 and Table 3
and 4, respectively.
Table 2 Sources of Data
Variables Source
FTTH
ADSL
CATV
Ministry of Internal Affairs and Communications
(http://www.johotsusintokei.soumu.go.jp/)
Price Reports of MIC
GDP
Data from National Accounts of Japan (SNA)
(http://www.esri.cao.go.jp/jp/sna/)
Intra-HHI, Inter-HHI Reports of MIC
Population
Data from Statistics Bureau and the Director-General for Policy Planning
(http://www.stat.go.jp/data/jinsui/)
CPI
Data from Statistics Bureau and the Director-General for Policy Planning
(http://www.stat.go.jp/data/cpi/)
Source: Author
3.2. Result of Estimation I: whole market
First, the estimation of the whole FTTH market (single house and condominium
markets together) is conducted, and the result is shown in Table 4. Let us summarize the
results as follows:
FTTH Price: West Japan (p<0.1) only satisfies the sign condition, and its price
elasticity is -1.79.
GDP: All Japan (p<0.01) and West Japan (p<0.01) satisfy the sign condition.
ADSL: This is positively significant in all cases (p<0.01), which shows the general
trend of migration from ADSL to FTTH.
CATV: All Japan satisfies the sign condition (p<0.05) indicating that FTTH is
substitute to CATV.
Intra-platform HHI: All Japan (p<0.01), East Japan (p<0.05), and West Japan
(p<0.1) are negatively significant. Competition among FTTH carriers promotes
diffusion.
Inter-platform HHI: All Japan (p<0.01), East Japan (p<0.01), and West Japan
(p<0.01) are negatively significant. The coefficient of East Japan Inter-HHI is the
9
largest. FTTH diffusion has been enhancing by competition among broadband
technologies such as ADSL and CATV.
Table 3 Summary Statistics (Eastern Japan)
Obs. Mean S.D. Min Max
All Japan
No. of subscribers: FTTH (Condominium+Single house) 1222 11.840 1.169 9.076 15.011
No. of subscribers: FTTH (Single house) 1222 11.506 1.041 8.821 14.003
No. of subscribers: FTTH (Condominium) 1222 10.449 1.488 6.242 14.556
Price (FTTH) 1222 8.589 0.056 8.393 8.742
Price (FTTH: Single house) 1222 13.307 0.022 13.250 13.395
Price (FTTH: Condominium) 1222 12.819 0.020 12.774 12.897
GDP 1222 17.760 0.030 17.695 17.827
No. of subscribers: ADSL 1222 11.954 0.885 10.217 14.564
No. of subscribers: Cable modem 1222 10.678 1.243 6.545 13.744
Intra-Platform HHI (FTTH) 1034 8.708 0.271 8.136 9.152
Intra-Platform HHI (FTTH: Single house) 1034 8.838 0.217 8.461 9.180
Intra-Platform HHI (FTTH: Condominium) 1034 8.519 0.348 7.689 9.158
Inter-Platform HHI 1222 8.380 0.152 8.115 8.899
East Japan
No. of subscribers: FTTH (Condominium+Single house) 442 12.168 1.257 9.226 15.011
No. of subscribers: FTTH (Single house) 442 11.755 1.097 8.936 14.003
No. of subscribers: FTTH (Condominium) 442 10.948 1.570 7.530 14.556
Price (FTTH) 442 8.576 0.067 8.393 8.712
Price (FTTH: Single house) 442 13.321 0.020 13.283 13.395
Price (FTTH: Condominium) 442 12.812 0.020 12.774 12.886
GDP 442 17.761 0.031 17.701 17.827
No. of subscribers: ADSL 442 12.373 0.873 10.764 14.564
No. of subscribers: Cable modem 442 10.641 1.571 6.545 13.744
Intra-Platform HHI (FTTH) 374 8.812 0.285 8.162 9.122
Intra-Platform HHI (FTTH: Single house) 374 8.977 0.179 8.658 9.180
Intra-Platform HHI (FTTH: Condominium) 374 8.602 0.346 7.756 8.993
Inter-Platform HHI 442 8.440 0.147 8.182 8.899
West Japan
No. of subscribers: FTTH (Condominium+Single house) 780 11.655 1.073 9.076 14.443
No. of subscribers: FTTH (Single house) 780 11.365 0.982 8.821 13.765
No. of subscribers: FTTH (Condominium) 780 10.166 1.361 6.242 13.734
Price (FTTH) 780 8.596 0.047 8.462 8.742
Price (FTTH: Single house) 780 13.299 0.019 13.250 13.373
Price (FTTH: Condominium) 780 12.823 0.019 12.774 12.897
GDP 780 17.759 0.030 17.695 17.825
No. of subscribers: ADSL 780 11.717 0.801 10.217 13.901
No. of subscribers: Cable modem 780 10.699 1.012 8.275 13.541
Intra-Platform HHI (FTTH) 660 8.649 0.244 8.136 9.152
Intra-Platform HHI (FTTH: Single house) 660 8.760 0.197 8.461 9.163
Intra-Platform HHI (FTTH: Condominium) 660 8.471 0.340 7.689 9.158
Inter-Platform HHI 780 8.346 0.145 8.115 8.794
Source: Author
10
Table 4 Result of estimation (Whole market)
Dependent variable: No. of subscriber
(FTTH): 2005Q3~2011Q4
All Japan East Japan West Japan
Price (FTTH)
-1.066 -1.307
-1.789*
[0.684] [1.179]
[0.925]
GDP
1.783***
0.477
2.409***
[0.345]
[0.581]
[0.424]
No. of subscribers (ADSL)
0.369*** 0.801*** 0.275***
[0.068] [0.133] [0.083]
No. of subscribers (CATV)
-0.018**
-0.014 -0.011
[0.009]
[0.016] [0.010]
Intra-Platform HHI
-0.347*** -0.314** -0.253*
[0.103] [0.157] [0.147]
Inter-Platform HHI
-1.028*** -1.348*** -0.690***
[0.105] [0.147] [0.163]
Constant
-14.889* 13.543 -23.605**
[8.590] [14.805] [11.215]
Observations
843 312 531
Number of ID (Prefecture)
47 17 30
R-squared
0.195 0.416 0.241
Standard errors in brackets
*** p<0.01, ** p<0.05, * p<0.1
3.3. Result of Estimation II: Single house market
Here the result of estimation on the market of single house, which is shown in
Table 6. The following variables are significant.
FTTH Price: East Japan only satisfies the sign condition (p<0.01), and its price
elasticity is -5.36.
GDP: All Japan (p<0.01) and West Japan (p<0.01) satisfy the sign condition.
ADSL: All Japan (p<0.01), East Japan (p<0.01), and West Japan (p<0.1) satisfy the
sign condition and they are positively significant. In particular, East Japan shows
larger migration from ADSL to FTTH.
CATV: All Japan (p<0.05) and West Japan (p<0.05) satisfy the sign condition
indicating that FTTH is competing more severely with CATV in West Japan. This is
consistent with the fact such that Western Japan contains prefectures with high
CATV diffusion rate such as Mie and Ohita.
11
Intra-platform HHI: All Japan (p<0.01) and East Japan (p<0.01) are negatively
significant. Since the Eastern Japan has larger share of single house in broadband
than the Western Japan, this fact causes the result.
Inter-platform HHI: All Japan (p<0.01), East Japan (p<0.01), and West Japan
(p<0.01) are negatively significant. The coefficient of East Japan Inter-HHI is the
largest. FTTH in the Eastern Japan, diffusion has been enhancing by competition
with other broadband technologies.
Table 5 Result of estimation (Single house)
Dependent variable: No. of subscriber
(FTTH): 2005Q3~2011Q4
All Japan East Japan West Japan
Price (FTTH)
-1.223
-5.364***
0.315
[1.050]
[1.997]
[1.207]
GDP
1.337***
0.828
1.497***
[0.363]
[0.646]
[0.427]
No. of subscriber (ADSL)
0.354*** 0.860*** 0.167*
[0.078] [0.157] [0.094]
No. of subscriber (CATV)
-0.020**
-0.003
-0.022**
[0.009]
[0.018]
[0.010]
Intra-Platform HHI
-0.462*** -0.698***
-0.212
[0.140] [0.248]
[0.173]
Inter-Platform HHI
-0.905*** -1.131*** -0.686***
[0.118] [0.194] [0.157]
Constant
-0.197 68.994** -27.87
[15.395] [29.256] [17.865]
Observations
843 312 531
Number of ID (Prefecture)
47 17 30
R-squared
0.185 0.308 0.13
Standard errors in brackets
*** p<0.01, ** p<0.05, * p<0.1
3.4. Result of Estimation III: Condominium
Finally, the result of estimation of the condominium market is shown in Table 6,
which can be summarized as follows:
FTTH Price: West Japan (p<0.01) only satisfies the sign condition, and its price
elasticity is -7.80, which is quite high. This is opposite result to the single house
market, in which East Japan shows higher price elasticity.
GDP: All Japan (p<0.05) and West Japan (p<0.01) satisfy the sign condition.
12
ADSL: All Japan (p<0.01), East Japan (p<0.01), and West Japan (p<0.01) satisfy the
sign condition and they are positively significant. In particular, East Japan shows
larger migration from ADSL to FTTH. No difference can be found between East
and West Japan.
CATV: All three variables are not significant.
Intra-platform HHI: All Japan (p<0.01), East Japan (p<0.01), and West Japan
(p<0.01) are negatively significant. There is no difference among three variables. In
the condominium market, competition among FTTH carriers is severe.
Inter-platform HHI: All Japan (p<0.05) and East Japan (p<0.01) are negatively
significant, while West Japan is not.
Table 6 Result of estimation (Condominium)
Dependent variable: No. of subscriber
(FTTH): 2005Q3~2011Q4
All Japan East Japan West Japan
Price (FTTH)
-6.786***
-1.485
-7.798**
[2.507]
[3.980]
[3.469]
GDP
2.556**
-0.472
4.396***
[1.009]
[1.325]
[1.397]
No. of subscriber (ADSL)
0.868*** 0.828*** 0.752***
[0.165] [0.312] [0.223]
No. of subscriber (CATV)
-0.003 -0.027 0.003
[0.025] [0.037] [0.033]
Intra-Platform HHI
-0.341*** -0.465* -0.316*
[0.132] [0.248] [0.168]
Inter-Platform HHI
-0.695** -1.009***
-0.623
[0.272] [0.385]
[0.415]
Constant
45.762 35.255 24.995
[34.362] [54.934] [46.673]
Observations
814 298 516
Number of ID (Prefecture)
47 17 30
R-squared
0.156 0.202 0.125
Standard errors in brackets
*** p<0.01, ** p<0.05, * p<0.1
3.5. Summary of analysis
The above three analyses show that Japanese FTTH as a whole promoted by
competition with other technologies such as ADSL and CATV, which can be seen as
migration to FTTH. In addition, completion among carriers in the FTTH market as well
as in the three broadband markets is high so that diffusion of FTTH is enhanced. In
13
comparison with East and West Japan, It seems that East Japan is more competitive than
West Japan. This is not consistent with the realty, and more elaborate estimation is
required. Table 7 shows the summary of the above estimations.
Table 7 Summary of estimations
Total Single House Condominium
All
Japan
East West
All
Japan
East West
All
Japan
East West
Price (FTTH) - - - -
GDP + + + + + +
No. of
subscribers
(ADSL)
+ + + + + + + + +
No. of
subscribers
(CATV)
- - -
Intra-Platform
HHI
- - - - - - - -
Inter-Platform
HHI
- - - - - - - -
4. DISCUSSION
In this section, we will compare the results of estimation with our previous papers
such as Tsuji and Akematsu [2010], Shinohara, Akematsu and Tsuji [2011] and [2012]
which analyzed in the same framework as this paper. In addition to this, other factors
which are not discussed in the empirical studies such as carriers’ managerial objectives
are discussed here.
4.1. Government policies and market conditions on FTTH
Unlike ADSL that use pre-laid metal subscriber lines, FTTH requires fiber-optic
subscriber lines. The Japanese government has started to work on a series of FTTH
diffusion policies which are the same as ADSL, namely openness of the fiber networks
such as unbundling and connection charges. Japan is only country which implemented
the open network policy, while other countries did not implement yet. According to
Shinohara, Akematsu, Tsuji [2012], the effect of unbundling to FTTH diffusion is found
to be negatively significant for OECD 30 counties, while unbundling to ADSL is
positively significant. Japan is thus an exceptional case.
One note, however, should be added in this context. NTT East and West, the
dominant operators, are obligated to open fiber-optic subscriber lines, just as it was with
14
ADSL service metal subscriber lines. This requirement, like the ADSL requirement, is
intended to establish unbundling, collocation, and connection fees so that other
operators are able to offer services that can compete with NTT East and West’s FTTH
services. However, as the technical attributes of fiber-optic technology limit unbundling
to a certain level, competitors that make use of NTT East and West’s fiber-optic
infrastructures cannot actually offer services that are completely competitive with those
of NTT East and West.
Since there are opposite opinions, it is difficult to decide whether Japanese
competition policy in the FTTH market promote its diffusion. Another factor will be
discussed later.
4.2. Business conditions and strategies of operators
The growing diffusion of FTTH and the phasing out of ADSL have become
prominent after NTT East and West, the owner of the metal lines, announced in
November 2004 its intentions to make a complete transition to the optical network and
the possibility of removing metal subscriber lines.
This move has created three major problems for competitors that had previously
focused on ADSL services. First, NTT’s announcement on the possibility of removing
metal subscriber lines made it increasingly difficult to concentrate primarily on DSL
and forced the providers to be more cautious about continued investment in
management resources. Second, NTT East and West’s full transition to FTTH would
mean a smaller ADSL market and higher speed broadband services in the entire market,
thereby putting pressure on competitors, as well, to move from ADSL to FTTH. Third,
competitors were unable to offer FTTH services which are sufficiently competitive with
NTT East and West, thus it was difficult to develop its broadband service operations. In
fact, Softbank, which had secured a share comparable to that of NTT East and West in
the ADSL market, abandoned its plans to make a full-scale entry into the FTTH market
(deciding not to provide FTTH as a Softbank service), and now sells NTT East and
West’s FTTH services as a wholesaler.
On the other hand, NTT East and West, the dominant operators, had the following
five major reasons that influenced their move:
(1) Since they are a strictly fixed (not mobile) operator, they intended to
concentrate on the enhancement of management resources including capital
investments in order to set FTTH as a mainstay business;
15
(2) They wanted to pool resources toward optical subscriber lines only, and avoid
the double burden of having both metal subscriber lines and fiber-optic
subscriber lines;
(3) They were locked in a battle with Softbank for ADSL share supremacy, with
both tied at roughly 35% (NTT East and West had provided all domestic
telecommunications services prior to the 1985 liberalization);
(4) They have suggested the possibility of removing metal subscriber lines, which
are vital to ADSL services. They had no intention to reinforce the permanency
of ADSL; and,
As described above, we could assume that the fierce competition between
operators in the ADSL arena, fueled by the provisions of broadband competition
policies that opened metal subscriber line infrastructures, prompted NTT East and West
to shift to FTTH. Then, having concentrated on management resources for FTTH and
made aggressive capital investment moves, NTT East and West has widened its lead
over competitors. It seems apparent that the decision of the dominant operator, who
owns the metal lines, on the transition of its business entirely toward the FTTH
subscriber network and the possibility of removing metal subscriber lines has caused
rapid FTTH diffusion in Japan and ADSL peak-out to a certain degree. Thus in the stage
of FTTH, NTT East and West, facing with heated competitor opposition that pulled its
share of the ADSL market down to a tie for the lead (35%), decided to make capital
investments in the FTTH subscriber lines, prompting the shift from ADSL to FTTH.
It should be noted carrier’s attitude toward FTTH businesses. In the development
of the DSL business, the amount of new investment is limited because the metal
subscriber line that had already been laid for telephones can be used without additional
modification. In contrast, the development of the FTTx business requires a vast
investment for laying optical fiber subscriber lines, which are indispensable for the
business. Furthermore, dominant carriers are generally considered to be negative toward
the provision of any new service that absorbs the profits of existing services. Such a
management attitude is called "cannibalism" (Falch [2007], Cambini and Jiang [2009]).
Thus, since DSL takes away the profits of existing services, such as from telephones
and dedicated lines, and FTTx does so with regard to DSL by providing a higher-speed
service, dominant carriers are said to be reluctant to provide these new services. It can
reasonably be said that the remarkable diffusion of FTTx in Japan is the result of
dominant carriers' management strategy to dare to accept cannibalism and the massive
16
capital investment mentioned above. In this way, they took a great risk. In contrast,
dominant carriers in Europe and the United States appear to have avoided such risk.
Such management strategies can be regarded as having affected the diffusion of FTTx.
5. CONCLUSION
In order to conclude this paper, we attempt to summarize policy frameworks
under discussions for FTTH. The issues to be addressed in this framework are as
follows:
(1) Connection charges
Currently interconnection charges of fixed telephone are based on LRIC
(Long-run Incremental Costs), while charges of dark fiber on expected future demand
and dry cupper on historical costs. NCCs insist on the charges of optical subscriber
network should be based on LRIC. In addition to access charges, there are other issues
to be solved such as unbundling of facility and the openness of platform of the security
and charging systems of providing application services via FTTH.
Under the current accounting system of NTT, costs attributed to FTTH are not
clearly itemized from the total costs, while the costs of interconnection of PSTN are
restrictedly calculated based on LRIC. Revenues related to FTTH are grasped by the
NTT branch offices, but FTTH costs are separated from the legacy network yet.
(2) Other regulation policy
The NGN network consists of different layers such as the access, platform and
application. In the access layer, the equal access to all subscriber lines such as CATV,
ADSL and FTTH. NTT’s NGN is based on openness to all possible networks which can
access to NGN. The platform function such as charging application services and
guaranteeing security must be separated from NGN, since NTN can manipulate this
function in order to exploit from application services. NGN also should be open to all
service providers. An example of the closed network is mobile telephone, which is a
vertical integrated business model in which all layers are owned by a mobile phone
company. Policy is required to how to secure the open NGN.
(3) Development of FTTH networks through competition
The development of ADSL in Japan was promoted by the following three factors:
17
namely, (i) deregulation such as collocation before June 2001; (ii) competition among
ADSL carriers from September 2001 to January 2002; and (iii) development of ADSL
technology which enabled high speed access such as 12 or 24 Mbps from October 2002
to January 2003 (Tsuji and Tomizuka, 2006; Akematsu, 2008a). These factors made
Japan as one of the most advanced broadband economy. This paper compared
development of FTTH with that of ADSL and clarifies the issues of the FTTH market.
Among these carriers, competition is based on “facility based.” In addition to this,
there is another type of competition, namely “service competition.” NCCs such as
KDDI and Soft Bank utilize NTT’s FTTH networks for providing their own broadband
services through the framework of “shared access.” The development of Japanese FTTH
has been promoted through these two types of competition. There are, however, still
problems to be solved. Even though the NTT’ share of total broadband services
including cable modem, ADSL, and FTTH is less than 50%, in the FTTH market their
shares are more than two-third. Japan needs delicate competition policy to make good
balance of two type competition. In order to promote more diffusion of FTTH, facility
competition is recommended, while in order to reduce NTT’s share service competition
is required.
(4) How to promote FTTH use
Recently one idea came out in order to raising ICT usage, being termed by
Broadband Super Highway Initiatives. Now 90% of Japanese can access to FTTH, but
only 30% actually utilize it. This initiative attempts to raise 90% to 100% and to use this
FTTH as a basis for various applications such as medicine, education, business, and so
on. In the area of medicine, for example, the Medical Law, established more than 100
years ago, prohibits telemedicine, since diagnosis or medical treatment must be
provided by face-to-face basis. In addition, telemedicine is not admitted by
reimbursement from public medical insurance. The attitude of Japanese people toward
privacy and security is another example. They refuse any national identify system being
afraid of government’s abuse. The unified ID system is a key to e-government as well as
e-commerce. The Japanese global ranking of e-government is rated quite low among
OECD counties.
Another alternative to enhance FTTH usage is the termination of PSTN, that is, if
the termination of PSTN is achieved, then the costs of its maintenances are transferred
to those of FTTH, and accordingly all subscribers will enable to use FTTH at 1,400 yen
(about US$ 15.00) per month and NCCs can use this network as their access network to
18
the Internet at low charges. With low charges, all people can enjoy FTTH. This idea
seems to be similar so-called “functional superstation” implemented in UK. The MIC
plans to achieve this in 2015. It is still the level of an idea, but the share of NTT locals
in the FTTH market, this kind idea becomes one possible alternative of competition
policy to utilize FTTH network.
REFERENCES
Akematsu, Y. [2008a] “A Panal Data Analysis of the Diffusion of ADSL Based on the
Data of ADSL Carriers,” Annual Report of Japanese Society of Information and
Communication Research 2007 (in Japanese), pp. 1-14.
Akematsu, Y. [2008b] “An Empirical Study of Japanese ADSL Development by Panel
Data Analysis on Four Major Carriers,” Proceedings of ITS Biennial Conference,
Montreal, Canada.
Akematsu, Y., and M. Tsuji [2007] “Deregulation or Market Competition, Which Has
Larger Effect on Japanese ADSL Development,” Proceedings of ITS European
Conference, Istanbul, Turkey.
Berkman Center [2010] "Next Generation Connectivity: A review of broadband Internet
transitions and policy from around the world".
http://cyber.law.harvard.edu/sites/cyber.law.harvard.edu/files/Berkman_Center_Broad
band_Final_Report_15Feb2010.pdf ((last accessed: 2 July 2010).
Cambini C., and Y., Jiang [2009] “Broadband investment and regulation: A literature
review,” Telecommunications Policy, Vol. 33, Issue 10-11, pp. 559-574.
Commission of the European Communities [2009] "Communication from the
Commission to the European Parliament, the Council, the European Economic and
Social Committee and the committee of the Regions Progress Report on the Single
European Electronic Communications Market 2008.
Directorate for Science, Technology and Industry, Committee for Information,
Computer and Communications policy [2009] "Working Party on Communication
Infrastructures and Services Policy, Indicators of Broadband of Broadband
Coverage".
FCC [2010] “High-Speed Services for Internet Access: Status as of Dec 31, 2008.”
Falch, M. [2007] “Penetration of broadband services – The role of policies,” Telematics
and Informatics, Vol. 24, No. 4, pp. 246-258.
Fiona, V. [2009] "World Broadband Statistics, Q3 2009," Point Topic ltd.
19
Kushida, K. [2009] “Political Economy of Telecommunication Policy - Comparative
Analysis of Japan, US and South Korea“ in T. Ida, S. Negishi, T. Hayashi (eds)
Studies on Information and Communications Policy (in Japanese), (NTT Publishing
Co., Ltd).
Ida, T., and T. Kuroda [2006] “Discrete Choice Analysis of Demand for Broadband in
Japan,” Journal of Regulatory Economics, Vol. 29, No. 1, pp. 5-22.
Kuroda, T. [2008] “An Empirical Analysis of the Indirect Network Effects in Japan's
Broadband Market,” A paper presented at the 2008 Japanese Economic Association
Autumn Meeting, Kinki University, Japan (in Japanese).
Lee, M. J. [2005] Micro-Econometrics for Policy, Program, and Treatment Effects,
Oxford University Press.
Ministry of Internal Affairs and Communications [2005] and [2007] Report of
Assessment of Competition in the Telecommunications Market 2004 and 2006, (in
Japanese).
Okada, Y., H. Ohashi, M. Noguchi, and M. Sunada [2006] “Demand Analysis of the
Broadband Market,” Competition Policy Research Center, Japan Fair Trade
Commission (in Japanese).
Perloff, J. M., L. S. Karp and A. Golan [2007] Estimating Market Power and Strategies,
Cambridge University Press.
Tanaka, T., K. Yazaki, and R. Murakami [2004] “An Investigation on Reality on
Competition in Broadband Services,” Competition Policy Research Center, Japan
Fair Trade Commission (in Japanese).
Shinohara, S., I. Sakaibara and M. Tsuji [2010] “Analysis of Deployment of Broadband
Services from the Viewpoints of Policy and Competition: Case of Japan and
International Comparison,” Proceedings of ITS European Conference, Copenhagen,
Denmark.
Shinohara, Sobee, Yuji Akematsu, Masatsugu Tsuji [2011], “What Promotes Migration
among Broadband Services in OECD 30 Countries: Panel Data Analysis,”
Proceedings of ITS Africa-Asia Regional Conference 2011, Perth, Australia.
Shinohara, Sobee, Yuji Akematsu, Masatsugu Tsuj [2012] “Empirical Analysis of
Factors Promoting Broadband Deploymant,” (in Japanese), Journal of Information
and Communications, Vol. 29. No. 4. pp.73-86.
Tsuji, M. and Y. Akematsu [2010] “New Policy Paradigm in the Age of NGN and
Broadband: Case of Japanese Telecommunications,” Proceedings of ITS Biennial
Conference, Tokyo, Japan.
Tsuji, M. and M. Tomizuka [2006] “An Empirical Analysis of Factors Promoting
Japanese Broadband: Case of ADSL,” Proceedings of ITS Biennial Conference,
20
Beijing, China.
Turner, S. D. [2006] "Broadband Penetration in the Member Nations of The OECD,
Why Does the U.S. Lag Behind? A comparative Statistical Analysis and the
Implications for Universal Service Reform," Free Press.
