An Exploratory Study on Coopetitive Behavior in Oil and Gas Distribution

Abstract

The purpose of this article is to analyze coopetitive behavior of companies operating in oil and gas distribution networks, enhancing both conceptual clarity of the concept and understanding of its specifics in oil and gas industry. We developed a model based on six factors: intensity, functionality, formalism, benefits, tension and stability to investigate 10 research hypothesis on a sample consisting of 154 subjects from 39 companies. By its conceptualization and results, our study is one of the first focusing on coopetitive behavior in oil and gas distribution and contributes to shape coopetition as a distinct subject for research.

1. Introduction

Discussion of coopetition, representing the simultaneous pursuit of cooperation and competition by organizations, has been on the rise in the academic literature recently [1,2,3,4,5]. Without any doubt, coopetition is a reality across many industries. Our paper seeks to contribute to the understanding of coopetition dynamics in oil and gas products distribution supply chains by empirically testing several factors specific to the coopetition process. By using a model which integrates various approaches of the concept, we analyze and discuss specific interactions and the resulting dynamics.

Scholars argue that coopetition between organizations create a paradox which have the potential to aggravate tensions among them and break existing partnerships [6,7]. Large failure rates of alliances between competitors [8,9] indicate that companies lack, in many cases, the ability to manage tensions occurring during coopetitive relations [10,11], negatively impacting the expected outcomes [12,13]. However, this does not mean that managing the coopetition paradox [14] is needless, requiring a balanced approach of emerging tensions [15,16,17] to fully grasp the benefits of cooperation.

Our research is based on a model relying on the literature dedicated to coopetition, covering the most important dimensions of any coopetitive relation: intensity, functionality, formalism, benefits, tension and stability. The outcome is coopetitive performance measured by subjects’ perception about overall benefits of getting involved in the coopetition. Our study is conducted in oil and gas distribution networks, a fertile ground in terms of coopetition considering the multitude of companies involved and the nature of their relationships. Romanian market is dominated by several large providers, while smaller companies distributing their products have to compete with other companies established by large oil companies themselves.

This paper aims to develop coopetition literature by clarifying the landscape of coopetitive relations in oil and gas distribution, enabling scholars and practitioners to better understand this process.

2. Literature Review

The engagement of organizations in coopetition materializes a paradox in the relationship [12,18]. Scholars argue that such a paradox engenders coopetitive tensions that in turn have the potential to aggravate the relationship and break partnerships [6,7], with large failure rates of alliances between competitors [8,9]. Unfortunately, the definition of coopetition still remains fuzzy [19]. Various definitions are used, with significant differences in terms of scope and nature. Usually, coopetition is defined as cooperation between directly competing organizations [5,20,21]. Moreover, coopetition has also been defined as an occurrence between different supply chains [22], and networks [23].

In strategic management literature, cooperation and competition were traditionally seen as antagonistic drivers within cooperative arrangements [24], with competitive side of a cooperative relationship usually regarded as potentially harmful [25,26]. By contrast, coopetition perspective tries to integrate the two paradoxical approaches into a common construct [14,27]. The emerging perspective is to depict cooperation and competition as two separate continua allowing to distinguish between different forms of coopetition with varying levels of intensity for cooperation and competition respectively [12,28,29].

According to Brandenburger and Nalebuff’s (1996) [30], this cooperative-competitive approach has to be defined relationally, as relations between customers, suppliers, and competitors, which jointly create value for themselves and the other partners involved. So, coopetition is perceived as the sum of various relations, with cooperative and competitive relations divided between various actors. This conceptualization is used by networking and industrial coopetition literature [4,31,32]. At an inter-organizational level, Bengtsson and Kock [33] emphasized the tension and complexity occurring when two or more organizations cooperate and compete simultaneously. They restrict coopetition at direct competitors, focusing on the horizontal coopetition and argue that relations are divided between activities. Hence, according to this approach, coopetition appears as consisting of two simultaneous interactions: cooperation for a distinct set of activities and competition for a different set of activities [1,3,5,11].

In the literature there are several approaches used by scholars when studying coopetition phenomena. One approach uses game theory, considering cooperative-competitive relations as mutual profitable relationships and discussing the balance between value creation and value appropriation [1,30,34]. A second approach uses a resource-based view and focuses on the benefits of mutually developing and using technologies and resources [35,36]. Finally, a third approach, based on networks, emphasize the importance of cooperative ties between competing firms in terms of learning and knowledge sharing through networks [37] or joint development of competitive advantages [20].

Other scholars develop the coopetition literature based on the level where cooperative-competitive relations occurs:

(a) On an individual level, coopetition relates to how individual motivations stimulate various individuals to compete, even if they trust each other [38]. The most frequently outcomes are knowledge sharing and team performance [39,40].

(b) On an organizational level, coopetition emphasize processes between units or teams, and primarily highlight how they affect knowledge sharing. Tsai [41] argues that formal hierarchy negatively affect knowledge sharing while informal lateral relations have positive effects on the process. For Luo [21] different infrastructural systems enable units to collaborate, while simultaneously competing with each other. Ghobadi and D’Ambra [2] analyzed cross-functional project teams and argue that the organization of projects is critical for positive outcomes in terms of knowledge sharing.

(c) The most common level of coopetition research is inter-organizational. Sometimes called simple coopetition, it occurs between organizations placed on the same level of value chain. This type of relational structure emphasize continuous and intensive transfer of products and information between supply chain participants, discouraging opportunistic behavior. According to Bengtsson et al. [19], partners seek increased competitiveness and securing competitive advantages, mutual development of technological innovations, exploration of international opportunities, and access to resources. Various studies explored how coopetition affects R&D activities [42], knowledge creation [43], knowledge acquisition [44], and knowledge transfer [45]. Commitment to long-term cooperation does not involve abandoning competition, the reason behind cooperation lies in the willingness to work with a vendor to solve technical and economic problems, instead of changing provider.

(d) Network-based coopetition, sometimes called complex coopetition, occurs in industrial clusters, where concentration of small and medium sized organizations in aggregated structures enable a dynamic network of closely interconnected structures, with varying levels of cooperation intensity. At the same time, network-based coopetition covers the entire value chain, from production to distribution. Scholars argue that organizational structures and processes serve as the backbone to balance coopetition, both at the organizational and inter-organizational levels; however, the latter level also demands the consideration of relation-specific and contextual factors as critical. Peng and Bourne [23] studied coopetition between healthcare networks and concluded that coopetition between networks with compatible but different structures is easier to balance. Song and Lee [22] studied how both cooperation and competition between value chains facilitate knowledge acquisition and the created logistics value. Studies on the network level coopetition are scarce, and more studies are needed in the future.

Distribution of oil and gas products is characterized by a highly complex network of companies that are involved in both cooperation and competition [46]. Hence, the oil and gas industry is a strong case to analyze coopetition because its products are widely used in economic and social activities [47] and the nature of interactions between various actors is dynamic, under continuous changes.

Coopetition assumes that companies involved facilitate strategic and transparent integration of key business processes and cooperation among supply chain members [48,49,50]. As such, coopetition entails that companies move beyond their immediate economic concerns and take proactive steps towards accepting and multiplying aggregated effects of cooperation while retaining their ability to compete on other markets or new clients. Srivastava [51] suggests that companies employ new strategies to achieve their goals. As such, propensity of companies from oil and gas distribution toward coopetition should help overcome negative effects of intense competition, a distinct feature in oil and gas distribution, while allowing them to sustain itself economically. Another study [52] argue that oil and gas companies could be exposed to legal, financial and reputational risks if they fail to align their objectives with the suppliers.

To address these risks, various scholars [53,54] have suggested actions that companies may take in order to improve supply chain performance. Improve cooperation with suppliers to improve the overall performance of logistics operations, for instance, may be useful in reducing costs and effective coordination of supply chain activities [55].

Involvement in coopetition in supply chains may have an internal and an external driver. Internal driver can be demonstrated through top management leadership [56,57,58,59,60], development of a more open organizational culture [48,49,57,61] or willingness to adhere to coopetitive business behavior. Management support is also crucial to enable mutual learning processes and to facilitate knowledge transfer [62,63]. A more open organizational culture determine companies to actively engage with suppliers and clients to improve supply chain performance, leading to vertical integration across supply chain and horizontal integration across networks [48,49].

Coopetition may lead to better risk management related to brand image and quality, enhancing the involved companies’ reputation [64], mitigating economic risks of supply chain activities [62]. Close cooperation with supply chain partners could facilitate more effective risk assessment practices [65,66]. As such, coopetition may enable collaboration between supply chain partners in developing supply chain solutions that can collectively improve their capabilities and competitiveness [67]. The lack of coordination and communication in companies and across supply chains could lead to ineffective cooperation [50,65]. Strategic collaboration with suppliers enables companies to develop supply chain capabilities and knowledge that can be used to reduce costs [59].

Oil and gas companies may find reasons to get involved in cooperation due to potential increase in perceived or expected benefits, surpassing benefits of acting alone or with traditional partners, and reduce threats, usually from external factors. Other benefits include better use of resources [2,68], better use of capabilities and skills to fulfill specific objectives as well as saving costs, both for current operations and for specific projects or increased organizational effectiveness. Others [69] argue that collaboration improve effectiveness because it enables companies to direct resources to increase their credibility, while for others [68] collaboration lead to a common framework to solve problems [70]. Other scholars highlight synergy achieved when two or more organizations combine their strengths [68] or improved delivery of services [70].

Concerning threats as a factor leading to cooperation, the literature includes elements like environmental turbulence [71] or political and economic circumstances [68]. According to this rationale, collaboration is a natural response to environmental factors. Increased competition for resources is such a factor [69]. Instead of competing for scarce resources, companies can team together, using less resources to get more value.

Regarding competitive dimension, namely barriers in active cooperation between companies, lack of motivation is perhaps the most common cited. It includes a wide array of factors, such as reluctance of many companies to consciously sacrifice something (knowledge, resources, etc.) in order to collaborate with others [72], or lack of organizational congruence between organizations, encompassing divergent goals or objectives [68], unwillingness to compromise their specific goals for the sake of common objectives [72], difficulties in managing obligations deriving from cooperation [70] or operational differences, relating to either day-to-day running of the business or a perceived incompatibility of leadership or management styles of the companies [72].

3. Materials and Methods

There is a significant lack of empirical research in the field, with most studies focusing on setting up models and frameworks. Hence, we investigated the coopetition phenomenon among companies operating in oil and gas distribution based on a quantitative research. Data collection was performed by using a standardized, closed questionnaire. The questionnaires were sent by email, with a small number filled directly during a workshop. In the end, we collected 154 valid questionnaires (

Table 1. Sample structure.

Surveyed Companies		Share in the Sample	Surveyed Respondents		Share in the Sample
Company size	Micro (<10)	5.19%	Respondent position in the company	President	5.19%
	Small (10–49)	38.31%		Executive	74.03%
	Medium (50–249)	56.49%		Board member	20.78%
Company age	<10 years	4.55%	Respondent education	ISCED 4 or less	0.00%
				ISCED 5 and 6	69.48%
	10–15 years	41.56%		ISCED 7 or more	30.52%
	>15 years old	53.90%	Respondent age	<35 years old	7.14%
Development Region	South-East and South	33.12%		35–50 years old	63.64%
	South-West and West	20.78%		>50 years old	29.22%
	Bucharest–Ilfov	30.52%
	North-East and North-West	11.03%

). The respondents were explicitly encouraged to note any clarifications they see fit to explain their answers. For the validity of our study, only respondents from top management tier were considered, like president, manager or board member were considered. This limitation was considered essential for acquiring data on coopetition is terms of trust or performance, for instance.

Capitalizing on literature on this topic [73] we defined a model composed of six factors (

Table 2. Conceptual model.

Variable	Description	Measurement Scale
Intensity factors
Number of partners	Number of companies with whom the company has established relations, regardless if cooperative or competitive.	Four intervals were defined: 1–3 partners, 3–10, 10–25 and more than 25 partners.
Behavior	The behavior of the surveyed company in the relation with others (both partners and competitors)	Two variables were defined: Proactive and Reactive. Binary type (0-No, 1-Yes)
Competitors	Number of companies which were considered direct or indirect competitors.	Two variables were defined: Direct competitors and Indirect competitors. Binary type (0-No, 1-Yes)
Number of competitors	Number of companies which were considered direct or indirect competitors.	Four intervals were defined: 1–3, 3–5, 5–10 and above 10 organizations.
Functionality factors
Value creation	Describes where coopetitive interaction is taking place in terms of value creation (value chain)	Possible values are: know-how transfer, resource sharing, operations, procurement, sales. Binary type (0-No, 1-Yes)
Initiator	Describes the initiator of the coopetitive relation	Two variables were defined: surveyed company or competitor. Binary type (0-No, 1-Yes)
Objectives	Describes the expected objectives to be achieved in the coopetitive relation	Possible values are: New services (provide new services), New markets (access new markets), Knowledge (access new knowledge from partners), Resources (access additional resources), Other (other objectives). Binary type (0-No, 1-Yes)
Structure	Measures the existence or not of a formal structure, responsible for coordination of coopetitive relation	Binary type variable (0-No, 1-Yes)
Formalism factors
Contract	Analyses the relation from a formal point of view, if there was any written contract/ arrangement between companies	Binary type variable (0-No, 1-Yes)
Trust	Seek to identify the opinion of the respondents about the relative importance of the contract in comparison with the trust in partner	Binary type variable (0-No, 1-Yes)
Trust evolution	Describes the evolution of trust between the partners after the start of the coopetitive relation	There were three possible answers: 1- Decreased, 2- Similar and 3- Increased
Benefits factors
Benefits	Describes the respondents’ perception about surveyed company benefits from coopetitive relation	Possible values: 1-Negative impact, 2-No impact, 3-Positive impact
Competitor benefits	Measures the perception of the respondents about the competitor’s benefits achieved by coopetitive relation	Possible values: 1-Negative impact, 2-No impact, 3-Positive impact
Benefits distribution	Measures the perception of the respondents about the distribution of mutual benefits.	Possible values: 1- More favorable for competitor, 2- Even distributed, 3- More favorable for surveyed company
Tension factors
Coopetitive tension	Measures level of tension during coopetitive relation	Likert scale from 1-Non-existent to 5-Very high
Opportunism	Describes the perceived risk of the opportunistic behavior from the partner	Binary Variable (0-No, 1-Yes)
Stability factors
Stability	Measures if the coopetitive relation is established on a determined or undetermined period	Binary variable (0-Undetermined; 1-Determined)
Performance	Describes overall perception of the respondents about the coopetitive relation	Possible values: 1-Negative impact, 2-No impact, 3-Positive impact
Future cooperation	Describes the willingness of the respondents to be engaged in similar relations in the future.	Three binary variables were considered: Yes, No, Undecided.

):

(1) Intensity factor reflects the number of traditional partners and collaborative competitors. A higher number of partners is considered to positively contribute to cooperation [74]. In terms of access to resources, a higher number of partners offers the chance to find additional resources and gain access to relevant know-how [68] while simultaneously signals better experience, superior skills in managing connections with other organizations and superior learning skills or knowledge retention capabilities. Hereby, we argue that H1.

Hypothesis 1 (H1).

Companies with a high number of partners achieve better results in coopetition.

(2) Functionality factor reflects how the coopetitive relation is working. Variables of interest are type of relation, highlighting to what level of the value chain occurs value creation, who was the initiator of the relation, what are the objectives pursued and if a separate structure for managing the relation was established [75]. Therefore, we argue that H2.

Hypothesis 2 (H2).

Value creation level influences benefits distribution.

(3) Formalism factor address the trust vs contract side of the coopetitive relationship, the existence of any agreements or contracts between partners, the mutual trust and how it has evolved during their relation. Two hypothesis were assumed, namely:

Hypothesis 3 (H3).

Distinct formal structures are required to manage successfully manage coopetitive relations.

Hypothesis 4 (H4).

Coopetitive relations based on trust are more beneficial than those based on formal agreements.

(4) Benefits factor considers company own benefits, competitors’ benefits and their distribution. Creating value and making use of it is critical for coopetition [1]. Companies can create shared value by collaborating with competitors, by accessing their resources or develop common sales, for instance. This value will be shared jointly by all the partners according to each one’s contribution and bargaining power. The problem of mutual vs private benefits has been addressed first by Khana et al. [72] and later developed by Kumar [76]:

Hypothesis 5 (H5).

Benefits are maximized when cooperation and competition are balanced.

Hypothesis 6 (H6).

Benefits of company are directly related to partners’ benefits.

(5) Tension factor was necessary to be included due to the paradoxical nature of coopetition, the necessity of balancing cooperation and competition, and the need to manage tension within the relation [77] to achieve a dynamic resource balance [78]. Scholars propose different strategies to manage this tension. Oliver [79] suggests that competition and cooperation should be separated for each process and in time, while Chin [80] emphasize the importance of trust. Tension is considered to have a beneficial effect on performance, but high levels are considered unproductive [81]:

Hypothesis 7 (H7).

Coopetitive tension is inversely proportional to coopetitive performance.

Hypothesis 8 (H8).

Trust is inversely proportional to coopetitive tension.

Hypothesis 9 (H9).

Increased opportunism corresponds to increased coopetitive tension.

(6) Stability factor reflects the impact of definite/indefinite character of relation or the stability and performance of coopetitive relations [82]:

Hypothesis 10 (H10).

Increased stability leads to low coopetitive tension and higher performance.

Hypothesis were correlated with factors of the model. Interactions between factors were further analyzed in a regression model to determine the variation in the level of benefits associated with the concurrent action of considered factors.

Special attention was payed to measurement of coopetitive relations Performance. There is no generally accepted method up to date. The selected method of measuring performance is subjective, perceived performance by top executives of surveyed companies. However, taking a look at the fact that decision-makers are engaged in a coopetitive relation for specific objectives, their individual perception of the effectiveness of the relation is relevant (possibly the most relevant) and other studies confirm the accuracy of the method [83].

The tools used in the analysis were non-parametric correlations (Kendall and Spearman), cross-tabs, analysis of variance and linear regression.

Before analyzing the actual data we tested their reliability with the common variation method (

Table 3. Common variation analysis.

	Initial Eigenvalues			Extraction Sums of Squared Loadings
Component	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	6.702	14.259	14.259	6.702	14.259	14.259
2	4.271	9.086	23.345	4.271	9.086	23.345
3	4.078	8.676	32.021	4.078	8.676	32.021
4	3.137	6.674	38.695	3.137	6.674	38.695
5	2.387	5.079	43.774	-	-	-
6	2.214	4.711	48.485	-	-	-
7	2.045	4.351	52.837	-	-	-
8	1.951	4.150	56.987	-	-	-
9	1.707	3.632	60.619	-	-	-
10	1.670	3.552	64.171	-	-	-
11	1.548	3.293	67.464	-	-	-
12	1.537	3.271	70.735	-	-	-
13	1.382	2.940	73.675	-	-	-
14	1.220	2.596	76.271	-	-	-
15	1.048	2.230	78.501	-	-	-
16	0.976	2.076	80.577	-	-	-
17	0.891	1.897	82.474	-	-	-
18	0.837	1.780	84.254	-	-	-
19	0.740	1.575	85.829	-	-	-
20	0.670	1.425	87.254	-	-	-

Extraction method: Principal Component Analysis.

). Using analysis of variance to analyze the effect of one factor only we extracted four factors explaining 39% of the total variance. The first is responsible for 14.259% of the variation, the second for 9.086% and the sum of the following two, 15.3%. Considering the results we concluded that the problem of common variation is not confirmed and we proceed further.

4. Results

(1) Intensity factor was analyzed to test Hypothesis H1.

(a) Number of coopetitive relations measured total number of interactions of the surveyed companies, regardless if it is with an organization perceived as competitor or not (

Table 4. Competitors and partners.

		Number of Partners ᶺ Number of Competitors Cross Tabulation
		Number of Competitors				Total
		1–3	3–10	10–25	>25
Number of partners	1–3	24	2	0	0	26
	3–10	26	16	4	2	48
	10–25	10	12	10	4	36
	>25	6	10	8	20	44
Total		66	40	22	26	154

). In our study, the median for the number of partners was 2.6, while for the number of competitors it was 2.

According to Park and Russo [74], a large number of relations/partners contribute to achieving good results. Gong et al. [68] argues that a high number of partners means better access to resources and know-how. A similar study conducted by Yamakawa [84] testing the impact of business alliances (they assumed that a large number of alliances indicates a high number of partners) achieved a weak and insignificant negative correlation (−0.04, n.a). It is possible that engaging in relations with a large number of organizations (partners and competitors) weaken companies’ relational capabilities or make them to pay less attention to each relation and, as such, relations are not optimized.

(b) Behavior describes who initiated the coopetitive relation and as such it may be proactive (initiated by surveyed companies) or reactive (initiated by a competitor). In the study, 70% of surveyed companies were found to be proactive, 9% reactive while 21% manifests both a proactive and reactive behavior according to circumstances.

(c) Competitors variable measured number of perceived direct and indirect competitors for surveyed companies. Direct or indirect competitor status is given by the type of the cooperative relations (horizontal or vertical). In our sample, half of surveyed companies use horizontally coopetitive relations, 16% use vertical relations while 34% were engaged in mixed relations, a balanced result considering the 2 main approaches in coopetition field: coopetition in a large sense [30,85], which includes mixed and vertical relations, and the narrow approach of coopetition [11], which considers only horizontal coopetition, relations with direct competitors.

We tested Hypothesis H1 which was not confirmed, the link was weak and statically insignificant (0.154, n.s.) (

Table 5. Correlation between Number of coopetitive relations and Benefits.

			Number of Coopetitive Relations	Benefits
Kendall’s tau_b	Number of coopetitive relations	Correlation Coefficient	1.000	0.154
		Sig. (2-tailed)	-	0.131
		N	154	154
	Benefits	Correlation Coefficient	0.154	1.000
		Sig. (2-tailed)	0.131	-
		N	154	154
Spearman’s rho	Number of coopetitive relations	Correlation Coefficient	1.000	0.175
		Sig. (2-tailed)	-	0.128
		N	154	154
	Benefits	Correlation Coefficient	0.175	1.000
		Sig. (2-tailed)	0.128	-
		N	154	154

).

(2) Functionality factor details how the coopetitive relation is working and seek to test Hypotheses H2 and H3. Again, four variables were constructed:

(a) Value creation variable pinpointed where coopetitive interaction generated value creation for the company. Reviewing literature, we opted out for Know-how, Resource sharing, Operations, Procurement and Sales.

(b) Initiator variable identified who initiated the relation. An interesting fact was that 16 of surveyed companies provided mixed answers, they were simultaneously initiators and partners. We consider that is normal taking into account the very nature of coopetition relation, a succession of coopetitive contingent games. This situation was theorized by Padula and Dagnino [5], by referring to the theory of positive sum game.

(c) Objectives variable measured what goals have to be achieved or what were the reasons to engage in a coopetitive relation. In our study they includes New products/services, New markets, Access additional resources and Access new knowledge, in this succession.

(d) Structure variable measured existence of any formal, distinct organizational arrangement, created exclusively for coordination of coopetitive relation. Theoretical models [75] argue that the existence of the separated structures have a positive impact on how effective is the coopetitive relation. In our study, we discovered a clear lack of formal structures, just 9% of respondents admitted using it. It seems that a majority of surveyed companies’ executives prefer to personally and informally handle relations with other organizations.

To test H2 hypothesis we used a matrix and calculated the Benefits median (

Table 6. Benefits median.

Level of Value Creation	Mean	N	Std. Deviation
Know-how	1.863	44	0.710
Resources	1.565	46	0.506
Operations	1.400	10	0.547
Procurement	2.000	12	0.632
Sales	1.619	42	0.669
Total	1.688	154	0.633

).

Important differences exists, the highest level being registered by those relations seeking shared Procurement (2) followed by Know-how (1.863) and Sales (1.619). Hypothesis H2 is hence confirmed. The importance of knowledge transfer in the coopetitive relations in various studies is confirmed in our study by the high level of associated perceived Benefits.

Hypothesis H3 is partially confirmed. Correlation between Structure and Performance was not confirmed (−0.1, n.s), while Structure clearly correlated with Number of coopetitive relations (0.215, p < 0.05) (

Table 7. Correlation between Structure and Number of coopetitive relations.

			Structure	Number of Coopetitive Relations
Kendall’s tau_b	Structure	Correlation Coefficient	1.000	0.215 *
		Sig. (2-tailed)	-	0.040
		N	154	154
	Number of coopetitive relations	Correlation Coefficient	0.215 *	1.000
		Sig. (2-tailed)	0.040	-
		N	154	154
Spearman’s rho	Structure	Correlation Coefficient	1.000	0.235 *
		Sig. (2-tailed)	-	0.040
		N	154	154
	Number of coopetitive relations	Correlation Coefficient	0.235 *	1.000
		Sig. (2-tailed)	0.040	-
		N	154	154

* Correlation is significant at the 0.05 level (2-tailed).

).

(3) Formalism factor describes the level of trust in the coopetitive relation and its evolution. Three variables were constructed:

(a) Contract measured the existence of a formal contract/arrangement between oil and gas distribution companies during coopetitive relation. In our study, the majority of companies (82%) has formal agreements for the ongoing of coopetitive relations. The main reasons were the clear definition of the mutual obligations and the guarantees in case of any litigations. The results are comparable with those of Schmoltzi and Wallenburg [73].

(b) Trust measured opinion of the respondents on the contract versus trust balance during coopetitive relation. 63% of respondents consider that contractual agreements are more important, 32% rely on trust while 5% consider them equally important. Less trust means more contract.

(c) Trust evolution measured evolution of trust between the partners after the start of the relation. In more than 50% of the cases it increased, while for less than 20% it decreased. We concluded that more oil and gas distribution companies begin to trust more their partners, regardless if they were competitors or not, during their cooperation.

To test H4 hypothesis we compared mean values for Contract and Trust variables related to Benefits.

We concluded that H4 hypothesis is confirmed, the relation based on trust having a mean (1.862) higher than those based on contract (1.596). The difference between the numbers of cases is given by the fact that some of the respondents gave an equal importance for the two of the possibilities (

Table 8. Benefits (Contract vs. Trust).

	Contract			Trust
	Mean	N	Std. Deviation	Mean	N	Std. Deviation
No	1.880	50	0.600	1.583	96	0.646
Yes	1.596	104	0.634	1.862	58	0.580
Total	1.688	154	0.633	1.688	154	0.633

).

(4) Benefits factor describes performance of coopetitive relation based on own-benefits vs mutual benefits. Kale et al. [83] have investigated different ways of measuring the performance in partnerships and consider that the executives’ opinion about the results is relevant and precise, they are the most suitable to assess if the relation fulfill organizational objectives. This factor was used to test Hypotheses 5 and 6.

(a) Benefits measured respondents’ perception about how lucrative is the coopetitive relation. According to our study, 75.3% of respondents consider that coopetitive relation results are positive, 14.3% without impact (the benefits are compensated by company contribution) and 10.3% consider that the result is negative. The results are very similar to those achieved by Duysters and de Mann [86], in their analysis on the partnerships’ effects on the performance.

(b) Competitor benefits measured the perception of the respondents about the competitor company benefits achieved by coopetitive relation. According to our study, more than 80% of respondents consider that coopetitive relations bring benefits to their competitors, also.

We concluded that both variables (Benefits and Competitor benefits) have a similar distribution, correlation is positive and significant (0.447; p < 0.001) according to the mutual benefits theory (

Table 9. Benefits-Competitor benefits correlations.

			Benefits	Competitor Benefits
Kendall’s tau_b	Benefits	Correlation Coefficient	1.000	0.447 **
		Sig. (2-tailed)	-	0.000
		N	154	154
	Competitor benefits	Correlation Coefficient	0.447 **	1.000
		Sig. (2-tailed)	0.000	-
		N	154	154
Spearman’s rho	Benefits	Correlation Coefficient	1.000	0.422 **
		Sig. (2-tailed)	-	0.000
		N	154	154
	Competitor benefits	Correlation Coefficient	0.422 **	1.000
		Sig. (2-tailed)	0.000	-
		N	154	154

** Correlation is significant at the 0.01 level (2-tailed).

). Hence, H5 was confirmed.

(c) Benefits distribution measured the perception of the respondents about the distribution of mutual benefits. A majority of the respondents, 75%, have considered it as fair and mutual. Corroborating the result with the coopetition typology (Gnyawali and Park [11]), we concluded that a balance between the competitive and collaborative sides of coopetition was reached. In the 25% other cases we encountered a dominant competitive dimension, with benefits appropriated by one partner or the other. In our study, the majority of the relations involved operational contingencies in procurement, resources and operations. Hence, coopetition gave the chance of accessing only some marginal opportunities which negatively affect the outcome. This way the H6 hypothesis was confirmed, similar to the Luo [28] and Gnyawali and Park [11] studies according to which performance is maximized when both cooperative and competitive sides of coopetition are balanced.

(5) Tension is a factor who recursively emerge in many studies about coopetition (Gnyawali and Park [87]). The paradoxical nature of coopetition determines a perceived tension for decision makers, the so called coopetitive tension. Achieving a high level of performance means keeping under control the coopetitive tension.

(a) Coopetitive tension measured the perceived tension during coopetitive relation. In our sample, the level of perceived tension (2.74 based on the Likert scale) was between low and medium.

(b) Opportunism variable measured the perceived risk of the opportunistic behavior from the other partner. The concept is derived from game theory. The opportunism is expressed by the risk that one of the two players involved in coopetition stop cooperating after it gets its desired resources or outcomes from the other player. In our study it has a distribution almost equal, 48% of respondents stating they have experienced opportunistic behavior risks from their partners.

Regarding Hypothesis H7, a majority of theoretical models considers that a high level of coopetitive tension harms the relation’s performance [81]. This is confirmed in our research, correlation between the two of the variables, Benefits and Coopetitive tension is negative (−0.435; p < 0.001) (

Table 10. Benefits—Coopetitive tension correlations.

			Benefits	Coopetitive Tension
Kendall’s tau_b	Benefits	Correlation Coefficient	1.000	−0.435 **
		Sig. (2-tailed)	-	0.000
		N	154	154
	Coopetitive tension	Correlation Coefficient	−0.435 **	1.000
		Sig. (2-tailed)	0.000	-
		N	154	154
Spearman’s rho	Benefits	Correlation Coefficient	1.000	−0.490 **
		Sig. (2-tailed)	-	0.000
		N	154	154
	Coopetitive tension	Correlation Coefficient	−0.490 **	1.000
		Sig. (2-tailed)	0.000	-
		N	154	154

** Correlation is significant at the 0.01 level (2-tailed).

).

We further tested the H8 hypothesis by using Kendall and Spearman test for Coopetitive tension with Trust evolution (

Table 11. Coopetitive tension-Trust evolution correlations

			Coopetitive Tension	Trust Evolution
Kendall’s tau_b	Coopetitive tension	Correlation Coefficient	1.000	−0.261 **
		Sig. (2-tailed)	-	0.008
		N	154	154
	Trust evolution	Correlation Coefficient	−0.261 **	1.000
		Sig. (2-tailed)	0.008	-
		N	154	154
Spearman’s rho	Coopetitive tension	Correlation Coefficient	1.000	−0.306 **
		Sig. (2-tailed)	-	0.007
		N	154	77
	Trust evolution	Correlation Coefficient	−0.306 **	1.000
		Sig. (2-tailed)	0.007	-
		N	154	154

** Correlation is significant at the 0.01 level (2-tailed).

).

H8 hypothesis was confirmed (−0.261; p < 0.01). Fang et al. [7] analyzed the relation between trust and tension and they achieved comparable results (−0.35, p < 0.001).

Finally, H9 hypothesis was tested by using Coopetitive tension and Opportunism variables (

Table 12. Coopetitive tension–Opportunism correlations.

			Coopetitive Tension	Opportunism
Kendall’s tau_b	Coopetitive tension	Correlation Coefficient	1.000	0.266 *
		Sig. (2-tailed)	-	0.010
		N	154	154
	Opportunism	Correlation Coefficient	0.266 *	1.000
		Sig. (2-tailed)	0.010	-
		N	154	154
Spearman’s rho	Coopetitive tension	Correlation Coefficient	1.000	0.294 **
		Sig. (2-tailed)	-	0.009
		N	154	154
	Opportunism	Correlation Coefficient	0.294 **	1.000
		Sig. (2-tailed)	0.009	-
		N	154	154

* Correlation is significant at the 0.05 level (2-tailed). ** Correlation is significant at the 0.01 level (2-tailed).

).

This hypothesis was also confirmed (0.266; p < 0.01). Fang et al. [7] have analyzed the tension-engagement relation and achieved a significant negative correlation (−0.29; p < 0.01). Assuming Engagement and Opportunism are opposite constructs, there seem to be a reversed correlation between them. Starting from this premise, our results are comparable with Fang (especially Spearman rho test).

(6) Stability factor, defined by two variables, reflects the impact of definite/ indefinite character of coopetitive relations on their stability and performance. Theoretical models argue that determined term relations will have higher levels of coopetitive tension, opportunism and a lower level of implication which will translate in inferior performance. This factor was used to test H10 hypothesis.

(a) Stability measured if the coopetitive relation is stable (established on a determined or undetermined period). According to our research, 24.61% of the respondents are involved in determined term relations while 39% in undetermined term relations.

(b) Performance measured overall perception of the respondents about coopetitive relations. In our study, Performance is highly correlated with Benefits (0.697, p < 0.001) (

Table 13. Correlation between Benefits and Performance.

			Performance	Benefits
Kendall’s tau_b	Performance	Correlation Coefficient	1.000	0.697 **
		Sig. (2-tailed)	-	0.000
		N	154	154
	Benefits	Correlation Coefficient	0.697 **	1.000
		Sig. (2-tailed)	0	-
		N	154	154
Spearman’s rho	Performance	Correlation Coefficient	1.000	0.718 **
		Sig. (2-tailed)	-	0.000
		N	154	154
	Benefits	Correlation Coefficient	0.718 **	1.000
		Sig. (2-tailed)	0.000	-
		N	154	154

** Correlation is significant at the 0.01 level (2-tailed).

). Duysters and Man [86] achieved similar results in their empiric study about the strategic alliances.

In this case, we tested H10 hypothesis and used a linear regressive model (

Table 14. The influence of Stability and Coopetitive tension on Performance.

Model Summary
Model	R	R Square	Adjusted R Square	Std. Error
1	0.547 a	0.299	0.280	0.5636

^a Predictors: (Constant), Coopetitive tension, Stability.

). Results indicates that Hypothesis H10 was confirmed (R² = 0.299, p < 0.05).

Stability related coefficient is negative because the variable is encoded 0 for undetermined and 1 for determined. Hence, if the value decreased, Performance related coefficient increased accordingly. Coopetitive tension is inversely proportional with Performance, as we demonstrated earlier and is positive correlated with Stability (0.241, p < 0.05) (

Table 15. ANOVA test results.

ANOVA a
Model		Sum of Squares	df	Mean Square	F	Sig.
1	Regression	10.026	2	5.013	15.781	0.000 b
	Residual	23.507	74	0.318	-	-
	Total	33.352	76	-	-	-

^a Dependent Variable: Benefits. ^b Predictors: (Constant), Coopetitive tension, Stability.

and

Table 16. Final results.

Coefficients a
Model	Unstandardized Coefficients		Standardized Coefficients	1	Sig.
	B	Std. Error	Beta
1 (Constant)	3.533	0.172	-	20.594	0.000
Stability	−0.326	0.136	−0.241	−2.405	0.019
Coopetitive tension	−0.250	0.057	−0.437	−4.369	0.000

^a Dependent Variable: Benefits.

). Fang et al. [7] have tested the correlation between cooperation and financial performance and also achieved a positive relation.

5. Conclusions

This study makes important contributions in terms of coopetition between companies in oil and gas distribution. In our opinion, coopetition as a new paradigm of organizational cooperation has reached a critical level necessary for shaping it as a distinct subject of research for scholars and practitioners from various industries, and oil and gas is no different. Even though the conceptualization of coopetition is still underway, empirical studies has to be done to better understand coopetition specifics.

Considering our study results, rejection of H1 hypothesis means that a company may engage in a limited number of coopetitive relations without fear of being less effective and, as such, cooperation with a high number of partners does not constitute a prerequisite to achieve better results or improved performance. A balanced approach in choosing partners and areas for cooperation, is desirable [88].

Confirmation of H2 hypothesis brings forward the reasons behind decision to engage in a coopetitive relation. Surveyed companies seek Procurement and Know-how instead of Resource sharing or Operations, a sign of a more strategic than operational approach. This concur the findings of other studies addressing knowledge and resource in energy sector [89], and may concur to innovation [90] or open innovation [91], particularly in networks.

Partial confirmation of H3 hypothesis means that companies have to be prepared to make use of coopetitive relations by creating distinct, formal structures to manage them, especially when dealing with a large number of partners. However, we do not find evidence an impact of the structure of overall coopetitive relations on performance.

Confirmation of H4 hypothesis means that most surveyed companies do not value formalism, choosing to trust their partners. As such, they neglect contractual agreements for loosely, informal cooperation based on mutual trust.

Confirmation of H5 and HG hypotheses links benefits with competition and cooperation. When companies are competing each other, the eventual cooperation for specific objectives seems less desirable, perhaps due to a concurrent lack of trust. Even though they cooperate, decision makers carefully assess their own-benefits and compare them to competitor’s results. This led to an equilibrium between cooperation and competitive dimensions, enabling performance to be maximized.

Confirmation of Hypotheses H7, H8 and H9 brings forward the understanding of coopetitive tension and opportunism on coopetitive on performance. In terms of coopetitive tension, this study provide similar results with other studies [7,87]. Hence, coopetitive tension is inversely proportional to performance, while trust is inversely proportional to coopetitive tension. Similarly, high levels of opportunism corresponds to a high coopetitive tension. Coupled with previous results, it show that benefits are maximized when cooperation and competition are balanced doubled by low levels of tension and opportunism.

Finally, confirmation of Hypothesis H10, stability (long terms vs short term relations) is very important for overall coopetitive relations performance, with determined term relations having higher levels of coopetitive tension and lower level of performance than undetermined ones.

In conclusion, our results confirmed other studies for most of the factor analyzed. This represents a proof of increasingly cooperative behaviors among companies in oil and gas distribution as a practical way to maximize performance and benefits not necessary from interactions with traditional partners, but also with competitors.

In terms of limitations, our study does consider the traditional benefits, like economic and organizational [92], while neglecting sustainable benefits, determined by ecological responsibility [93] or use of new IT&C technologies in fostering cooperation [94].