The Importance of Government Support for Pipeline Network Construction

Abstract

This chapter describes that, focusing on economics theory, a country needs its government support to construct a pipeline network throughout the country. In Japan, before deregulation, vertical integrated companies (gas utilities) provided natural gas to customers in their monopoly area respectively. When the companies transport gas into their own areas, the companies choose to construct pipelines or to use LNG tank trucks from the sight of their strategies. Focusing on long-term uncertainty, short-term uncertainty, and locations, this chapter analyzed the factors of pipeline constructions. The results indicate that if there is another gas utility near a company, then the company construct a pipeline to the gas utility to transport gas. In contrast, if there are no neighbor utilities, the company tends to purchase gas via LNG tank truck. This means gas companies do not construct a pipeline network, or do not try to do it, but construct point to point pipelines. Therefore, without government supports, a pipeline network would not be constructed throughout the country.

1. Introduction

This chapter considers the importance of government support for pipeline network construction focusing on Japan’s natural gas industry from the perspective of economics, in particular, transaction cost economics and the organizational forms of gas local distribution utilities.

1.1 Overview of natural gas import

Before proceeding to outline this study, this section provides an overview of the domestic natural gas supply chain. Nearly all of Japan’s natural gas requirements have been imported from overseas via LNG tankers. In 2020, 92.02% of all-natural gas was imported as LNG, 4.36% was produced from gas domestic fields, and the remainder was generated from imported petroleum-based gas (Agency for Natural Resources and Energy, Gas Market Division, 2021). Tokyo Gas and Tokyo Electric Power Company (TEPCO) began to import liquefied natural gas (LNG) in 1969. Since then, increasing gas consumption has resulted in increasing imports of LNG. Figure 1 shows domestical natural gas use, and indicates that approximately 60% of LNG is consumed for power generation, and “Town Gas” is basically used as cooking and heating by end users.

Figure 1.

Natural gas use in Japan (source: Ministry of Economics, trade, and industry).

In recent years, many companies have started to import LNG, for example, the three major incumbents, Tokyo Gas, Osaka Gas, Toho Gas, and following incumbents with LNG terminals (Saibu Gas, Shizuoka Gas, and Hokkaido Gas). In addition, both upstream companies and power generation companies also import LNG. Here, the term ‘upstream company’ refers to a company that specializes in the production and transportation of energy, such as Japan Petroleum Exploration Company Limited (JAPEX) and INPEX Corporation, while power generation companies include Tokyo Electric Power Corporation (TEPCO), Kansai Electric Power Corporation (KEPCO), and Chubu Electric Power Corporation (CEPCO).

1.2 Retail markets and pipeline networks

The gas retail market has two main features. First, the companies are classified into two categories based on ownership structure. As of March 2021, 173 utilities were private companies while 20 were municipality-owned companies. Of the 173 privately owned utilities, 12 were listed companies. Second, the size of these firms varies significantly. As can be seen in Table 1, the maximum revenue is USD 12,586,010,000 (Tokyo Gas), while the minimum is USD 294,740. The large and medium-sized incumbents are involved in production (import), transmission, and distribution, that is, they are vertically integrated utilities. Meanwhile, there are medium- and small-sized incumbents that only sell gas to consumers. These incumbents purchase natural gas from upstream companies via pipelines or LNG tank trucks.

	Revenue	Production	Customers	Tangible Assets
(Unit)	USD	1000 MJ	People	USD
Average	171,474,800	8,421,867	188,521	135,625,510
S.D.	1,052,895,090	55,953,815	1,037,167	731,440,770
Min.	294,740	5590	510	239,650
Max	12,586,010,000	691,882,206	12,208,885	8,989,040,000

Table 1.

Basic information relating to gas distribution utilities (2015).

(S.D.: Standard Deviation, Source: Gas Business Annual Report, 2015)

(1USD = 100JPY)

One of the reasons for this situation is that a regional monopoly policy has previously been enforced in the natural gas industry. As far as we can see the natural gas industry after 1945, gas utilities had never been integrated politically, whereas electric utilities were integrated into 10 groups. Also, the spread of LPG use (liquified petroleum gas) since 1945 had greatly influenced the establishment of small sized natural gas utilities. Therefore, the natural gas industry has many utilities (incumbents) and large difference of the largest and smallest utilities. This policy has also affected the characteristics of the natural gas distribution network. Figure 2, in which the trunk pipeline networks¹ are depicted, shows that the trunk pipeline networks are quite poor, that is, the coverage area of networks is narrow and they are not enough connected, although the natural gas consumption in 2016 was 111.2 billion m³ which represented about 3.1% of global consumption, and Japan is the largest LNG importing country in the world².

Figure 2.

Pipeline networks and LNG terminals in 2016 (source: The Japan gas association).

As is well known, LNG is transformed into natural gas by regasification facilities at or close to LNG terminals. Each large incumbent typically constructed LNG terminals at a sea port close to cities with large populations in its monopolistic supply area and constructed new pipelines after estimating the profits that would be generated by the additional investment. Meanwhile, middle or small incumbents purchase LNG via tank truck or natural gas via pipelines from upstream companies. In the former case, incumbents constructed gasified facilities, and in the latter case, incumbents constructed trunk pipelines to neighbor suppliers. According to the Gas Business Annual Report (2015), total transportation volumes via pipelines and tank trucks were approximately 1735 billion MJ and 1324 billion MJ, respectively, in 2015. Even Tokyo Gas, the largest gas company, uses LNG tank trucks to haul natural gas in a part of its supply area. Consequently, pipeline networks radiated outwards from the 35 existing LNG terminals, and becomes narrow networks. Hence, the pipeline networks terminate in each region, and there are insufficient trunk pipelines connecting the various regions. Incumbents with a vertically integrated structure would not have sufficient incentive to connect their trunk pipelines with those of other incumbents. In fact, there was no pipeline connecting Tokyo and Osaka, a total distance of approximately 2450 km, until December 2015 (Source: Gas Business Annual Report, 2015).

Figure 3 shows the LNG terminals and pipeline networks of Osaka Gas, with the supply area shaded red. It is easy to see that high-pressure (red) and medium-pressure (green) pipelines spread outward from the two LNG terminals. Figure 3 also shows that the Japanese pipeline network is sparse. The supply area for Gojo Gas is shaded yellow. Although Gojo city in the Nara prefecture is located approximately 41.8 km southeast of Osaka city, there are no pipelines connecting the Osaka Gas and Gojo Gas.

Figure 3.

Osaka gas pipeline network and LNG terminals (source: Ministry of Economics, trade, and industry).

Sadorsky (2001) [1] indicates that it is difficult to introduce product differentiation in relation to natural gas, therefore, gas suppliers (utilities) are most likely to face price competition. Weir (1999) [2] describes that if an incumbent owns both transport and distribution facilities, it would be a barrier to entry for entrants that only have third-party access to pipelines. Hence, the UK government introduced unbundling regulations to increase the number of new entrants and improve competition in the UK retail market.

1.3 Natural gas supply chains

The natural gas supply chain varies according to their historical and geographical characteristics.

Sailer et al. (2009) [3] define a natural gas supply chain as following six stages: exploration, extraction, production, transportation, storage, and distribution. In this study, the supply chain is simplified three stages: import, gasification, and distribution (Figure 4). The “import” activities include transportation from overseas, exploration, extraction, and production because it focuses on the process of delivering imported gas to end users. “gasification” activities involve procedures to gasify LNG into natural gas and to transport gas from upstream companies to distribution utilities. “distribution” activities include both storage and distribution into end users. In general, Upstream companies operate “import” activities while downstream companies operate “distribution” activities. Regarding “gasification” activities, in some cases, the upstreams operate, and in the other cases the downstreams do. Basically, no transportation companies with pipelines or regasification facilities that are independent of the upstreams and downstreams exist in Japan (Figure 5). Thus, either an upstream or a downstream company needs to shoulder the responsibility for regasification activities to re-gasify LNG and transportation.

Figure 4.

The natural gas supply chain.

Figure 5.

Organizational structures.

Unlike in the United States and EU countries, the Japanese Government (regulatory authority) had never enforced unbundling regulations that prohibit management of both transportation (including import) and distribution activities until 2020, but since April of 2022, the government has introduced the regulations into three largest companies. Besides, almost all upstream companies basically own gas storage, while local distribution utilities (downstream companies) have to undertake the responsibility for stable gas supply to end users. Also, the Government had authorized local distribution utilities to provide natural gas on the principle of a natural monopoly, permitting the utility a business license in the permitted area. These utilities are obligated to provide natural gas to end users in their own area safely and continually, however, the government had not imposed any relevant regulations on gasification activities.

Many upstream companies have both gasification facilities and trunk pipelines to provide into local distribution utilities with natural gas through pipelines or with LNG using tank trucks (Figure 5). Because raw commodities such as natural gas is impossible with product differentiation, the best performing natural resource companies are generally the lowest cost producers (Sadorsky 2001) [1]. Therefore, taking into account cost minimization and management strategies, the utilities purchase natural gas directly by joining a pipeline from its own facility to trunk pipelines owned by upstream companies³, or purchase LNG by tank trucks. If distribution utilities purchase LNG directly, then they need to construct in-house gasification facilities to provide natural gas into end users. As a result, some downstream utilities have in-house gasification facilities, while the others do not (Figure 5).

1.4 Government policy and utility’s decision

A utility’s decision on whether or not it needs to establish gasification facilities is critical to that utility’s attempts to manage its economic performance. Besides, three historical or crucial circumstances might affect this decision: the Integrated Gas Family 21 plan (IGF 21) issued by the Ministry of International Trade and Industry in 1990⁴, official network plans by the government, and managerial uncertainties.

Next, regarding pipeline network plans, the government has not yet made official plans for a pipeline network, nor has it provided financial aid for its construction, although both retail prices and supply areas have been regulated for a long time. When incumbents implement construction of their own pipelines, they need to raise the money for its construction, and, the decision to implement it depends upon long-term demand and managerial efficiencies. In the case where a massive amount of gas is transported, pipelines are superior to LNG tank truck, but, the former option requires huge capital expenditure to build pipeline facilities. When an incumbent encounters large uncertainties related to the weather conditions (meteorological conditions) or a volatile industrial demand that is affected by the economic conditions, it tends to refrain from an investment in a trunk pipeline even if large demand is expected. In these cases, the incumbent selects to purchase LNG via tank trucks. Thus, the government has never been strongly concerned with pipeline construction.

Also, managerial uncertainties might affect the vertical integration choice. Stable procurement is an indispensable part of distribution utilities. However, it might be difficult to implement an obligation to sustain security of supply for a long time. This is because even if managerial uncertainties are large, the distribution utilities have to continue providing natural gas in constant and sufficient quantities for a long time. To decrease the uncertainties, some utilities strive to purchase gas from plural wholesalers to maintain multiple supply chains, while others set up multiple natural gas storage tanks. Hence, pipeline construction would be affected by political issues, uncertainties, stable procurement. As a manner to explore pipeline construction factors, this study, focusing on a transaction cost economics theory, estimates the transaction cost empirically, and then considers the importance of comprehensive construction policies.

2. Theoretical background

2.1 Transaction cost economics

Here, the context of Transaction Cost Economics (TCE) is defined.

Coase (1937) [4] predicted existence in external costs between two firms, and for a single firm, internal costs exist between its divisions. The concept of external and internal costs was considered to be one of the significant factors when an entrepreneur determines a firm’s boundaries. When engaging in business transactions, a firm has a strong incentive to integrate with another firm that has significantly high external costs. In contrast, if the external costs between the two firms are not very high, then the former firm does not have a strong incentive to integrate with the latter firm, though would continue to do business with it.

The external and internal cost concept defined by Coase (1937) [4] was developed into transaction cost economics by Williamson (1975, 1985, 1995) [5, 6, 7], who defined the term “invisible costs” as “transaction costs”, and explained the origin of transaction costs based on three factors: (a) uncertainty, (b) relationship-specific assets, and (c) frequency (Williamson, 1985) [6].

There are several notable papers based on transaction cost views of vertical integration in the manufacturing industry. First, Monteverde and Teece (1982) [8] analyze asset specificity of GM and Ford, and found that the probability of vertical integration between a parent company and a subsidiary might rise because the manufacture of parts which needs advanced technologies in automobiles tends to become relationship specific assets. Second, Masten et al. (1989) [9], separating specific assets into human assets and physical assets, insist that human assets affect vertical integration more than physical assets. Third, Walker and Weber (1984,1987) [10, 11] show that, focusing on the uncertainty, when the uncertainty to get manufacturing parts becomes higher the probability for vertical integration also becomes higher.

Regarding empirical analyses, Levy (1985) [12] estimates the boundaries of firm by using 67firms’ data (37industries), and puts asset specificity as R&D investment, and moreover puts uncertainty as variance of sales. As a seminal work for transaction cost economics of power generation industry, Joskow (1985, 1988) [13, 14] found that power generation plants incline to be constructed close to mining pits, and that vertical integration between plants and pits, and long-term contracts, were widely practiced. Crocker and Masten (1996) [15] investigated the organizational forms of public utilities in the United States.

Shelanski and Klein (1995) [16] reviewed many empirical literatures related to transaction cost economics theory, and then concluded that they have conspicuously consistent with predictions from the theory. In contrast, David and Han (2004) [17] and Carter and Hodgson (2006) [18], from traditional literature surveys, found that asset specificity and uncertainty have received considerable scrutiny or commonly examined, whereas frequency has not. Hence, they concluded that some literatures have produced results that the transaction cost framework would not predict.

Sheravani et al. (2007) [19] noted the importance of the relationship between transaction costs and market power. They suggested that high market power appears to provide safeguards to a firm using nonintegrated channels not envisioned in predictions from transaction cost economics. Furthermore, they argued that firms with high market power are likely to have significant monitoring and surveillance capabilities, can exercise legitimate authority, and offer diverse incentives to associated channel members⁵.

2.2 Organizations and LNG supply chains

This section introduces several related literatures on LNG supply chains. Xunpeng, (2016) [20] points out that almost all the incumbent gas companies in Asian have vertically integrated supply chains. Lee et al. (1999) [21] found that KOGAS (the Korean national firm) have a lower productivity level compared to firms with acquiring their gas through pipelines because it depends on LNG import, which requires additional capital facilities for shipping, storage, and regasification. As been described by Vivoda (2014a) [22], a number of international LNG trade are dominated by long-term contracts. This is because the trading companies need the huge capital costs including liquefaction and regasification facilities and the inherent inflexibility in the value chain required contractual arrangements to protect both the suppliers and the buyers. In contrast, Cabalu (2010) [23] and Hartley (2013) [24] described that technological innovations make LNG transportation costs decrease significantly and LNG import and export volume were gradually increasing. Also, Gkonis and Psaraftis (2009) [25] indicate that LNG shipping markets are basically oligopolistic, and then suggested that competing companies have to consider a transportation capacity in the LNG shipping market. Vivoda (2014b) [26] points out that it is important for Japan and Korea to elaborate diverse LNG strategies to import LNG smoothly.

Turner and Johnson (2017) [27] denote that LNG trade is superior to pipeline transportation, and then point out that importers and exporters can easily send and receive gas to any locations with liquefaction and regasification facilities when LNG trade is possible. Xunpeng (2016) [20] and Hashimoto (2020) [28] describe the prospect in Asian LNG spot and hub markets.

3. Methodology and data

Here, this section describes the application of transaction cost economics theory into gas utilities. According to this theory, transaction costs consist of external and invisible costs between two firms, while internal costs are composed from invisible costs between two divisions in a single firm (Figure 6). A firm basically assumes to determine the choice whether or not it selects vertical integration, based on the transaction cost economics theory. Figure 6 illustrates the application of transaction cost economics. When transaction costs exceed internal costs, then firm B would merge with firm A or acquire it. This study defines this type of consolidation as a vertical integration. In contrast, when internal costs exceed transaction costs, firm B would not merge with firm A.

Figure 6.

Schematic of application of transaction costs and internal costs.

Regarding empirical estimations, this theory has two prudent treatments. First, Williamson (1985, p.20) [6] indicates the existence of ex ante and ex post transaction costs, and Monteverde and Teece (1982) [8] investigated ex ante transaction costs. This study also attempts to investigate ex ante transaction costs. Second, although both transaction costs and internal costs should be observed directly, it would be impossible to measure internal costs precisely. However, because transaction cost economics theory expects a positive correlation between transaction costs and incentive for vertical integration this study estimates only transaction costs to analyze the incentive.

Williamson (1985) [6] explains uncertainty, relationship-specific assets, and frequency as transaction costs. Many empirical analyses treat with uncertainty and relationship-specific assets [8, 9, 10, 11, 12]. Frequency is not analyzed because of the difficulty in estimation. This study also adopts the two components; uncertainty and relationship-specific assets. Regarding uncertainty, this study divides it into long-term and short-term components for its importance to the natural gas industry. In distribution utilities, because growth rate and demand fluctuation would be the main components of uncertainty, this study defined sales volume, number of customers, and average revenue growth rate as sources of long-term uncertainty. Also, Gas demand and underpinning sales would be affected by seasonal factors, which are defined as monthly sales variance and the inventory rate. These are components of short-term uncertainty in this analysis⁶.

Meanwhile, relationship-specific assets are classified into the assets by site specificity, physical asset specificity, human asset specificity, and dedicated assets by Williamson (1985) [6]. This study employs physical and site specificities, and does not analyze physical assets, human assets, and dedicated assets. For physical asset specificity, gas utilities comprise public and private administrations. Public gas utilities are expected to receive aid from municipalities when they face bankruptcy. Therefore, they may decrease transaction costs for related to relationship-specific assets.

To scrutinize long-term uncertainty, short-term uncertainty, and site specificity, the Integration equation is assumed by means of the methodology of Levy (1985) [12] and Wang and Mogi (2017) [29];

$$\mathit{\text{Integration}}=f\left(\mathit{LU},\mathit{SU},\mathit{SS}\right)$$E1

where LU, SU, and SS are long-term uncertainty, short-term uncertainty, and site specificity, respectively. If transaction costs increase, the value of Integration would become high. Hence, the high value means high incentive to integrate.

The dependent variable, “vertical integration (VI)”, represents whether or not a utility has gasification facilities. When a distribution utility has gasification facilities, it means that transaction costs exceed internal costs. Hence, “vertical integration (VI)” is adopted as the dependent variable. The dependent variable (x) represents whether or not a utility has gasification facilities.

Table 2 shows the definitions of dependent and independent variables, including the expected sign. RGR, SAL, CUS, ASS, and PRO are defined as components of long-term uncertainty, and SDR, SDM, SVV, AVI, and HHR are defined as components of short-term uncertainty. SSD and PUD are adopted as components of site specificity. The data sources are gas business annual reports, and SSD was obtained from the natural gas supply area map (agency for natural resources and energy).

Variable	Definition	Expected sign
VI	If the utility possesses regasification facilities, then assign 1. Otherwise assign 0.
RGR	Average revenue growth rate from 2006 to 2015 (absolute value)	+
PRO(LU)	Production volume in 2015 (1000 MJ)	+
ASS(LU)	Utility’s tangible assets in 2015 (1000 JPY)	+
SAL(LU)	Utility’s sales volume in 2015 (1000 MJ)	+
CUS(LU)	Number of customers in utility’s monopoly area (People)	+
SVV(SU)	Standard deviation of utility’s monthly sales volume from January to December divided by the whole sales volume in 2015; $\left(\frac{\text{Monthly sales volume variance from January to December in}2015}{\text{The sales volume in}2015}\right)$	+
SDM(SU)	Standard deviation of utility’s monthly sales volume from January to December	+
AVI(SU)	Average natural gas inventory for the past 3 years (absolute value) $\left(\frac{\text{Product}-\text{Sales}}{\text{Product}}\right)$	+
HHR(SU)	Household rate $\left(\frac{\text{Household sales volume}}{\text{Wholesales volume}}\right)$	—
SDR(SU)	Standard deviation of revenue for the 10-year period (2006–2015)	—
SSD(SS)	Site specificity dummy (2015); If a utility borders other distribution utilities or there are domestic natural gas fields in its own area, assign 1. Otherwise assign 0.	—
PUD(SS)	Public utility dummy (Public utility: 1, Private utility: 0)	—

Table 2.

Definition of variables (source; Hashimoto, 2021 [30]).

In this study, the integration eq. (1) was defined as:

$$\begin{array}{c}{e}^x=\alpha \prod _l{\mathit{LU}}_l^{b_l}·\prod _m{\mathit{SU}}_m^{c_m}·\prod _n{\mathit{SS}}_n^{d_n}\\ ln{e}^x=ln\alpha \prod _l{\mathit{LU}}_l^{b_l}·\prod _m{\mathit{SU}}_m^{c_m}·\prod _n{\mathit{SS}}_n^{d_n}\\ x=\beta +\sum _l{b}_{l}ln{\mathit{LU}}_l+\sum _m{c}_{m}ln{\mathit{SU}}_m+\sum _n{d}_{n}ln{\mathit{SS}}_n,\end{array}$$E2

where x means the probability of integration. LUl, SUm, and SSn respectively mean the l-th long-term uncertainty, m-th short-term uncertainty, and n-th site specificity, while α and β are constants. The high value of x means high incentive to integrate, while the low value means low incentive. Table 3 shows the descriptive statistics.

	Integration	RGR	PRO	ASS	SAL	CUS	SVV	SDM
average	0.624	0.043	8,462,922	13,534,961	7,445,264	189,432	0.018	87,643
S.D.	0.034	0.007	3,917,333	5,121,035	3,355,212	72,610	0.001	37,816
Min.	0	0.000	6451	23,965	6436	510	0.003	148
Max	1	1.015	691,882,206	898,904,000	577,580,996	12,208,885	0.043	6,516,900

Table 3.

Descriptive statistics (source; Hashimoto, 2021 [30]).

(S.D.: Standard deviation)

4. Results

Table 4 shows the Probit model results. The results of model 1–6 indicate robustness. Because strong correlations among independent variables of long-term uncertainty (LU) might be shown, these variables were not applied simultaneously⁷.

	Model 1	Model 2	Model 3	Model 4	Model 5	Model 6
Constant	2.656* (1.54)	1.923 (1.56)	2.622* (1.54)	2.396 (1.46)	2.369 (1.52)	2.622* (1.54)
RGR(LU)	−0.190* (0.10)	−0.189* (0.10)	−0.190* (0.10)	−0.186* (0.10)	−0.202* (0.11)	−0.190* (0.10)
PRO(LU)	0.142 (0.06)
ASS(LU)		0.086 (0.07)
SAL(LU)			0.017 (0.06)
CUS(LU)				0.071 (0.06)
SVV(SU)	0.474* (0.26)	0.555** (0.26)	0.478* (0.26)	0.533** (0.26)	0.517** (0.26)
AVI(SU)	0.197*** (0.07)	0.192*** (0.69)	0.197*** (0.70)	0.186*** (0.07)	0.191*** (0.07)	0.197*** (0.07)
HHR(SU)	−0.132 (0.70)	−0.120 (1.82)	−0.129 (0.19)	−0.140 (0.18)	−0.103 (0.19)	−0.129 (0.19)
SDR(SU)					0.049 (0.06)
SDM(SU)						0.017 (0.06)
SSD(SS)	−0.930*** (0.26)	−1.016*** (0.22)	−0.936*** (0.23)	−0.978*** (0.21)	−0.979*** (0.22)	−0.936*** (0.25)
PUD(SS)	−0.487* (0.30)	−0.514** (0.30)	−0.487* (0.30)	−0.463 (0.30)	−0.481 (0.30)	−0.487** (0.29)
R-squared	0.208	0.214	0.208	0.212	0.210	0.207
Log likelihood	−119.409	−118.551	−119.395	−118.868	−119.140	−119.395
Observations	205	205	205	205	205	205
Franction of collect predictions	0.68	0.70	0.68	0.69	0.69	0.68

Table 4.

Probit model results (source; Hashimoto, 2021 [30]).

(Note: Standard errors are in parentheses, and ***, **, and * are significant at 1, 5, and 10%, respectively)

The three types of transaction costs will be discussed⁸. First, for long-term uncertainty, the coefficients of PRO, ASS, SAL, and CUS are not significant at 10% level. RGR is significant at 10% in all models, but the sign was negative, not being consistent with the expected sign. Hence, the component of long-term uncertainty is required careful interpretation⁹.

Second, for short-term uncertainty, AVI was significant at 1% in all models, and standard deviation of SVV was also significant at 5% or 10%. HHR, SDR and SDM were not significant at the 10% level. Regarding coefficients for AVI, the consideration for a causal relation is required carefully because it would be possible to interpret that utilities with possessing regasification facilities increase their inventory volume. This study concluded that short-term uncertainty cannot be strongly supported.

Third, the coefficients of SSD were significant at the 1% level in all models, and also, those of PUD were significant at 5% or 10% level in some models. Therefore, the existence of assets and site specificities are strongly recognized.

5. Conclusions

This chapter, to confirm the importance of comprehensive pipeline network plans by the government, examined whether or not local distribution utilities integrated gasification activities, in terms of transaction cost economics theory, and then found that the utilities prefer to purchase natural gas via pipelines when there are wholesalers or neighboring utilities that provide gas to end users, or there are natural gas fields in the vicinity. As a result of the utility’s behavior, a broad pipeline network would not be built throughout the country.

Pipeline construction depending on distribution utilities affects the growth of broad pipeline networks. Because a utility prefers to construct a point-to-point pipeline between itself and a neighbor wholesaler the utility by no means construct a pipeline network by itself. Even if a number of utilities invest the point-to-point pipelines, a nationwide pipeline network cannot be built. Consequently, as seen in Figure 2, a broad pipeline network infrastructure has not been constructed throughout Japan. Therefore, this study concluded that we need the government support for pipeline network construction.

Table 5 shows the cases of pipeline investment. Table 5 indicates that companies invest comparatively small distance pipelines that are under 100 km, and moreover, joint ownership is adopted by the two related firms such as distribution utilities and wholesalers¹⁰. The former would reinforce this study that distribution utilities tend to purchase gas from those companies when there are wholesalers or neighbor utilities in the vicinity. Meanwhile, the latter might bear out the necessity of government supports for nationwide pipeline networks. The government basically has not supported the companies by means of pipeline network planning, comprehensive infrastructure policies, or financial support. The reason why joint-ownership is adopted might be without any governmental supports. In other words, this might be indicating that the government need to have some supports to build a nationwide pipeline network throughout the country.

No.	Pipeline’s name	Organizational form (Equity rate)	Year	Distance (Bore)
1	Ogijima-Kawasaki	Kawasaki Gas Pipeline Co. Ltd. {Tokyo Gas: 50%, Nihon Oil: 50%}	2004	5 km; (400 ∼ 500 mm)
2	Shizuoka Line; Showa-Gotenba	INPEX. Co. Ltd.	2006	81 km; (400 mm)
3	Minamifuji Line; Fuji-Gotenba	Minamifuji Pipeline Co. Ltd.; {INPEX, Shizuoka Gas, Tokyo Gas: 33% respectively}	2006	31 km; (500 mm)
4	Setouchi Line; Mizushima-Fukuyama	Setouchi Pipeline Co. Ltd.; {Hiroshima Gas:80%,Fukuyama Gas:20%}	2006	40 km; (300 mm)
5	Koriyama Line; Shiroishi-Koriyama	INPEX co.:80%,Tohoku Electric Power:20%; (Joint Ownership)	2007	95 km; (400 mm)
6	Shizuhama Line; Shizuoka-Hamamatsu	Shizuhama Pipeline Co. Ltd.(part of 76 km); {Shizuoka Gas: 50%, Chubu Gas: 50%}	2012	113 km; (400 ∼ 500 mm)
7	Chiba-Kashima Line	Tokyo Gas	2012	79 km; (600 mm)
8	Isewan Odan Pipeline; Chita,Kawagoe-Yokkaichi	Chubu Electric Power:50% Toho Gas: 50%; (Joint Ownership) Undersea tunnel	2013	19 km; (600 ∼ 700 mm)
9	Mie-Shiga Line; Hikone-Yokkaichi	Osaka Gas(20 km), Chubu Electric Power(40 km)	2014	60 km; (600 mm)
10	Himeji-Okayama-Line	Osaka Gas Co. Ltd.	2014	86 m; (600 mm)
11	Toyama Line; Itoigawa-Toyama	INPEX. Co. Ltd.	2014	102 km; (500 mm)
12	Saito Line	Tokyo Gas	2015	40 km; (600 mm)
13	Ibaragi-Tochigi Line	Tokyo Gas	2015	81 km; (600 mm)
14	Furukawa-Maoka Line	Tokyo Gas	2018	50 km; (600 mm)
15	Ibaragi Line	Tokyo Gas	2021	92 km; (600 mm)
16	Amagasaki-Kugayama Line	Osaka Gas	2028	49 km; (600 mm)

Table 5.

The cases of pipeline investment.

In addition, the government decided to unbundle natural gas companies. The three largest incumbents (Tokyo gas, Osaka gas, and Toho Gas) were unbundled in April, 2022. While the unbundling regulation is expected to promote competition, it might also discourage pipeline investment. Hence, incentives for investment after the introduction of unbundling regulation might need to be considered.

Finally, this study explored the determinants of vertical integration and the importance of pipeline network plans. However, this study has not considered or investigated the characteristics of network externalities, a natural monopoly, economies of scale, and management strategies that cannot be evaluated by transaction cost economics¹¹. They will be the focus on future work.