Skip to main content

The Behavioral Effects of Competition Intensity and Cost Structure on Competing Suppliers: An Experimental Study in the Context of the USA


This research investigates how competition intensity and differences in cost structures affect decisions made by competing suppliers and the role that behavioral factors play as influences. We use controlled laboratory experiments to study the scenario of suppliers competing for a share of demand being outsourced by a single buyer. The buyer seeks to maximize the service level provided by suppliers by allocating based on different performance measures which create varying levels of competition intensity. The experimental treatments include those performance measures as well as differences in supplier cost structures. Our experimental results show that in the majority of cases suppliers’ decisions do not confirm theoretical predictions from the Nash equilibrium, and we find patterns in those deviations. To explain them, we first evaluate behavioral factors found in the literature including bounded rationality, learning, and other-regarding behavior. We then introduce a new behavioral factor, rival-chasing. Rival-chasing builds on other-regarding behavior by considering competitors’ actions in addition to their outcomes. We find that rival-chasing can explain patterns in suppliers’ behavior that cannot be explained by other behavioral factors.


Competition plays a crucial role in the design of outsourcing mechanisms (Li, 2013; Li & Wan, 2017; Xu et al., 2016). It can lead to a higher capability and performance of the firms to have lower prices, higher quality, and better innovation (Kumar, 2021). Thus, it is important to have vigorous competitions that are designed based on appropriate business practices (Gundlach et al., 2019) to improve firms’ supply chain and operational performance. Supply chain is one of the management functions having an important effect on business competitiveness (Sonar et al., 2020). For example, Toyota, Honda, and Volvo frequently encourage competition between their suppliers and award the contract to the suppliers with the best performance (Dubois & Fredriksson, 2008; Liker & Choi, 2004). Nike and Cisco have multiple competing suppliers in their network and evaluate their performance against one another to allocate demand and to improve the business performance (Li, 2013).

There are few studies of competitiveness from the supply chain perspective, which makes it a great research opportunity in this area (Momaya, 2020). In this study, we aim to understand how design of a supply chain affects its performance by investigating the behavior and decisions of its players in an outsourcing competition. Based on the literature, we refer to the ability of a competition to induce players to exert more efforts as competition intensity and experimentally study how it impacts the competing suppliers’ decisions. Additionally, we examine how heterogeneity in suppliers’ costs influences their decision-making behaviors. We are interested in investigating how competition intensity and cost heterogeneity affect competing suppliers’ other-regarding preferences, rationality, and learning process. Moreover, we introduce a new behavioral factor, ‘rival-chasing’, and model it to explain the observed behaviors.

The findings from our study improve our understanding of the role that behavioral and cognitive factors play in the decision-making process. Our findings also provide insights into the interplay of competing organizations’ decisions and how businesses respond to other players in a supply chain. Understanding the decision-making mechanism in supply chain competitions can help managers in designing better sourcing mechanisms. It also creates practical insights into sourcing and supplier selection decisions, which play critical roles in purchasing and supply management (Giunipero et al., 2019; Wetzstein et al., 2019). Moreover, our findings shed more light on our knowledge of firms’ behaviors under different competitive environments, which is important in developing organizational capabilities and enhancing competitiveness (Shee et al., 2011).

The remainder of this paper is organized as follows. In the next section, we present the related literature to this study. Then, we explain the theoretical model of the supply chain competition. Afterward, we state our hypotheses, which is followed by our experimental design in the next section. Then, we present the results of our experiments and evaluate the factors that explain the observed behaviors in the following section. Finally, we offer our concluding remarks.

Literature Review

In the supply chain literature, outsourcing through competition has been widely studied from the theoretical standpoint (Dasci & Guler, 2019; Heese et al., 2020; Li et al., 2020; Zhao et al., 2014). The literature shows the way a competition is set up greatly affects players’ incentive to exert higher effort in their attempts to earn higher payoffs (Anderson & Freeborn, 2010; Benjaafar et al., 2007; Cachon & Zhang, 2007; Elahi, 2013). However, the question of how the design of the competition setup, and consequently the competition intensity, affects the suppliers’ decisions has not been well studied, particularly from an experimental standpoint. The literature shows that human decisions deviate from normative models in most cases, and experimental studies help us in testing the models empirically to find new theoretical and practical implications (Cannella et al., 2019; Shinde et al., 2020).

There are relatively few experimental studies of simultaneous competitions in supply chains. Of these, Chen et al. (2012) and Chen and Zhao (2015) are among the first. These studies model the competition between retailers in a capacity allocation problem and show that bounded rationality can explain retailers’ order behavior. In a similar setup, Cui and Zhang (2017) develop a behavioral model based on cognitive hierarchy theory to explain the retailers’ decisions in a capacity allocation game. They compare bounded rationality and cognitive hierarchy models and find out their relative abilities in explaining decision-makers’ behavior depend on problem context and parameter values.

A few studies examine behaviors of the competing newsvendors. These studies employ a combination of logit choice model and mental accounting (Yingshuai Zhao & Zhao, 2016), a strategic experience-weighted attraction model (Feng & Zhang, 2017), loss aversion (Villa & Castañeda, 2018), reference dependency (Villa & Castañeda, 2018), mean anchoring (Villa & Castañeda, 2018; Yukun Zhao & Zhao, 2017), and over-estimation (Yukun Zhao & Zhao, 2017) to explain competing newsvendors’ decisions. Ovchinnikov et al. (2015) develop a mathematical model to provide the best response to a behaviorally biased newsvendor in a competition.

In the context of competition between suppliers who face a single buyer, Hu et al. (2017) examine suppliers’ capacity investment behavior and find that bounded rationality and relative standing of the suppliers explain, to some extent, the observed capacity over-investment. A detailed review of these studies can be found in Donohue et al. (2018) and Fahimnia et al. (2019). The above studies investigate supply chain players’ decisions in a competition; however, none of them examine how competition intensity impacts those decisions. Table A1.1 in the ESM Appendix presents a summary of these relevant studies.

Moreover, the literature shows that non-identical players in a competition make different decisions, and players who have some advantages (e.g., lower costs) over others can benefit from this situation to gain more (Li et al., 2015). In an outsourcing competition, the supplier with a better performance, which can result in a lower cost or higher quality, provides a higher service level or offers a lower price to the buyer (Jain et al., 2020). We test this finding in our setup and examine how decisions of the more efficient suppliers with lower costs are different from the decisions of the suppliers with higher costs.

This study also contributes to the literature of other-regarding behavior (Cooper & Kagel, 2016) and dynamic competitive response (Jain & Hazra, 2017; Jin et al., 2019). These studies examine how players make their decisions in interaction with one another and show that social and outcome-based preferences affect players’ decisions. Katok and Pavlov (2013), Chen et al. (2015), Du et al. (2015), Chen et al. (2017), and Johnsen et al. (2019) show how other-regarding behavior in form of fairness preferences (inequity aversion) plays a significant role in the players’ decisions and consequently supply chain performance. We investigate decision-makers’ other-regarding behavior in more detail and shed light on the interplay between competitors’ decisions by examining effects of the competition intensity on them. Moreover, we study other-regarding behavior by examining how subjects’ behaviors is affected not only by their competitors’ payoffs (outcomes), but also by their competitors’ decisions (actions).

Problem Formulation

Our model consists of a buyer and N potential suppliers, in which, the buyer intends to outsource the production of a physical good to her supply base. The suppliers operate in a make-to-stock production fashion with a base-stock inventory policy. The buyer’s demand follows a Poisson process with a rate of (Benjaafar et al., 2007; Cachon & Zhang, 2007). The portion of demand randomly allocated to supplier i is δi, assuming 0 < δi < 1 and \(\sum\nolimits_{i = 1}^{N} {\delta_{i} } = 1\). Consequently, the share of each supplier from the buyer’s demand will be δi .

The suppliers adjust their capacity such that their production time, which is assumed to follow an exponential probability distribution with a mean of µi, maintain a targeted utilization δi. The utilization of supplier i is defined as ρi = δii, with the constraint of 0 < δi < 1. As a result, the production system at each supplier follows an M/M/1 queuing system.

Having a base-stock inventory policy at supplier i means that upon the arrival of one unit of demand, to maintain the targeted base-stock level zi (zi > 0), a replenishment order is sent to the supplier’s production system. The service level that each supplier provides for the buyer can be measured in terms of fill-rate, si, which is the probability of meeting the buyer’s demand from on-hand inventory. That is, si = Pr(Ii > 0) in which Ii is the on-hand inventory at supplier i. It is not difficult to verify that, for an M/M/1 queuing system, \(s_{i} = 1 - \rho_{i}^{{z_{i} }}\), which means the service level is an increasing function of base-stock inventory level.

We assume suppliers backorder the demand that cannot be satisfied from on-hand inventory while the buyer is willing to wait for the backordered demand. We also exclude the possibilities of the buyer acquiring the product from another source outside of the current pool of suppliers or switching to another supplier when one supplier is out of stock. Supplier i can provide higher service levels for the buyer by stocking higher inventory levels and incurs a holding cost hi per unit of inventory per unit time. Moreover, each supplier incurs a production cost ci per unit produced, and a capacity cost ki per unit of capacity, with the latter measured in terms of the rate of production.

The selling price per unit of product, p, is assumed to be the same across all suppliers. This happens when, for example, the buyer is powerful enough to set the price or the price is set by other market mechanisms. Acquiring the product from all suppliers at the same price, the buyer is then interested in inducing suppliers to provide higher service levels by intensifying the competition. To achieve this goal, the buyer orchestrates a competition by setting a demand allocation rule based on which each supplier can increase his portion of demand by increasing his committed inventory level compared to other suppliers. To choose the inventory level, each supplier tries to maximize his expected profit by considering the trade-off between increasing his demand share through keeping more inventory versus lowering the inventory holding cost by keeping less inventory.

In our setup, the choice of demand allocation rule affects the intensity of the competition and hence the inventory level that suppliers guarantee to keep (Elahi, 2013). We consider three types of allocation rule, which allocate the buyer’s demand to suppliers proportional to three different performance measures: (1) fill-rate, (2) base-stock inventory level, and (3) a combined measure that includes both fill-rate and base-stock inventory, designed to maximize the intensity of the competition to its highest level. ESM Appendix A2 shows the mathematical relations that govern these three types of competition designs and the corresponding equilibrium points. Elahi (2013) shows that, among these three competition designs, the competition based on fill-rate creates the lowest level of competition intensity and hence the lowest inventory level at the equilibrium point, while the competition based on the combined performance measure creates the highest competition intensity and hence the highest inventory level at the equilibrium point. Therefore, in this paper, we name the three abovementioned competitions, Low-Intensity, Medium-Intensity, and High-Intensity competitions, respectively.

Hypothesis Development

Elahi (2013) theoretically studies how the choice of the performance measure affects suppliers’ decisions in an outsourcing competition. The author shows that suppliers provide higher inventory levels for the buyer as the competition intensity increases. The author also shows that in a competition with non-identical players, low-cost suppliers provide higher inventory levels for the buyer compared to the suppliers with higher costs. Although these theoretical findings demonstrate how competition intensity and cost heterogeneity affect the decisions in an outsourcing competition, the vast literature of behavioral operations management shows that human decisions deviate from the results of the normative models. To test whether behaviors in practice are consistent with the theoretical findings on suppliers’ decisions in the setup of this study, we develop our hypotheses and examine them experimentally.

H1: In an outsourcing through competition setup, the inventory (and hence service) level that suppliers provide for the buyer increases as the competition intensity increases.

H2: In an outsourcing through competition setup, suppliers with lower cost structures commit to higher inventory (and hence service) levels.

Experimental Design

We conducted a series of experiments in which two suppliers compete for the demand share of a buyer (N = 2). Our experiments consist of six different treatments with three treatments apiece for different competition intensity levels. For each of these intensity levels, one treatment was for suppliers with identical costs and one for suppliers with heterogeneous production costs.

For all experiments, the price of the product, p, is 100, and the arrival rate of demand at the buyer is = 1.7. Suppliers adjust their capacities to maintain a utilization of = 0.93 and incur a capacity cost of k = 5 and an inventory holding cost of h = 1. For suppliers with identical cost structures, production costs are the same at c1 = c2 = 20. For suppliers with heterogeneous costs, the values c1 = 20 and c2 = 60 are used. We choose a relatively large difference between higher and lower costs so that any impact arising from the cost heterogeneity can be observed more evidently.

In our experiments, subjects played the role of competing suppliers, each having the goal of maximizing their profits. Experiments consisted of 30 independent rounds in which subjects needed to make their decisions on the committed inventory level in competing with one competitor. To run the experiments, we used a between-subject design and randomly assigned each treatment to the students of a class section at the College of Management at a state university in the US who have taken a business course related to supply chain management (Cui & Zhang, 2017; Hu et al., 2017). Each subject participated in only one treatment, and we ended up with a total of 115 subjects: 31 in Low-Intensity, 34 in Medium-Intensity, and 50 in High-Intensity competitions. We provided cash prizes as incentive for subjects to focus on maximizing their profits and awarded subjects in accordance with their achieved total profit at the end of each experiment session.

A strict protocol was followed for conducting all experiments, which began by asking subjects to read a two-page handout describing the supply chain setup, the decision-making process, and the competition itself. A short demonstration of the software to be used was then presented. The experiment software provided an interactive calculation tool available to the subjects throughout the experiments, as have other researchers used similar decision support tools for their experiments (Hu et al., 2017; Kalkanci et al., 2011). This calculation tool let the subjects focus less on calculating values and more on decision-making and evaluating results. Subjects, to maximize profit, needed to consider that the proportion of demand they receive depends not only on their own decisions but also on their competitors’ decisions. The tool enabled our subjects to evaluate any prospective decision on a pro-forma basis by displaying their profits as a function of the range of decisions their competitor might make. A sample screenshot of the software for experiments is shown in ESM Appendix A3 (Figure A3.1). After subjects became familiar with the software and gained a sense of how their decisions, combined with potential decisions by competitors, affects their profits, we answered any remaining questions.

The next step in the protocol was to run a 5-round practice competition to ensure subjects understood the whole competition setup clearly. We answered any remaining questions and then started the experiment.

Each round ended after either (a) all subjects had each entered a decision, or (b) the time limit had expired. At that point the software paired subjects randomly as anonymous competitors, with any subject not making a decision unpaired and assigned a profit of zero. For treatments with heterogeneous cost suppliers the software randomly assigned subjects to two groups. One group of subjects had the higher cost for c (production cost) and the other group the lower cost. Subjects were apprised of their group and made aware of the profit implications. We explained to the subjects that their profits would be normalized with respect to their costs at the end of the experiments to assure prize money be calculated fairly for all subjects.

After each round was completed, each subject was shown the decisions and profits in the round made by both the subject and their competitor, alongside the subject’s market share in the round and his cumulative profit to that point. Figure A4.1 in ESM Appendix A4 presents a flowchart of the experiment stages. On average, each session lasted around one hour.


Figure 1 shows the results of our experiments. The graphs in this figure depicts the subjects’ average decisions (base-stock level) over each of the 30 decision rounds of each experiment, as well as the corresponding Nash equilibrium.

Fig. 1

Diagrams of the subjects’ decisions in comparison with Nash equilibrium. Low-intensity competitions: a identical suppliers; b non-identical suppliers. Medium-intensity competitions: c identical suppliers; d non-identical suppliers. High-intensity competitions: e identical suppliers; f non-identical suppliers

Table 1 shows a summary of these results which includes the average base-stock decisions made by subjects over all rounds. We use the Wilcoxon Rank-Sum test to verify significant differences (Bolton & Katok, 2008; Cui & Zhang, 2017; Katok & Pavlov, 2013) and conduct our analyses based on the subjects’ average decisions over each round (Chen & Zhao, 2015; Chen et al., 2012; Hu et al., 2017) with R.

Table 1 Comparison of base-stock decisions under different competition designs

Competition Intensity and Service Level

Table 1 compares decisions from experimental results for different competition designs with different competition intensities under different cost structures. The results show that subjects’ average base-stock level, and hence the associated service level, increases with the competition intensity level. Subjects’ average decisions under High- and Medium-Intensity competitions are significantly (p < 0.001) greater than the average decisions under Medium- and Low-Intensity competitions, respectively. This holds true for both suppliers with identical and non-identical costs. Therefore, our results support hypothesis H1.

Cost Heterogeneity and Service Level

The results in Table 1 for the cases that suppliers’ cost structures are heterogeneous show that the subjects’ average decisions are higher when suppliers have lower costs. Our analysis shows the increase in suppliers’ decisions due to lower cost is statistically significant (p < 0.05). So, the experimental results support hypothesis H2.

Suppliers’ Decisions and Nash Equilibrium

Table 2 compares the experimental results with the corresponding theoretical Nash equilibrium for identical and non-identical cost suppliers.

Table 2 Nash equilibrium vs. experimental results base-stock levels

We observe in Table 2 and Fig. 1 that subjects’ behaviors vary under different competition designs. It is interesting to note that the average decisions in Low- and Medium-Intensity competitions are higher than the corresponding Nash equilibrium; however, subjects make decisions that their average is lower than the corresponding equilibrium points in High-Intensity competitions. Our analysis shows these differences between the experimental results and theoretical predictions are significant (p < 0.01) in most cases. We also observe different patterns of deviations with respect to high-cost and low-cost suppliers. ESM Appendix A5 presents further analysis of the subjects’ decisions.

Behavioral Factors

To model the possible underlying behavioral factors that result in the observed deviations, we use an extended version of the Quantal Response Equilibrium (QRE). In this model, we consider (a) random errors, as a source of bounded rationality, (b) learning, as a means of explaining the change in subjects’ decisions over successive rounds, and (c) other-regarding preferences, as a mechanism that explains subjects’ interactions and how the competitors’ profit affects subjects’ decisions. We then introduce (d) rival-chasing and examine if it adds to explaining subjects’ unexplained behaviors after the others are considered.

One of the popular approaches used in the literature to explain subjects’ behavior is based on the assumption that decision-makers aim to make the best decision but make random errors due to their bounded rationality (Johnsen et al., 2019; Ren & Huang, 2018; Song & Zhao, 2017; Song et al., 2020). For factor (a) above we use the concept of QRE to model randomness in players’ decisions (McKelvey & Palfrey, 1995). Following what we discussed in the experimental design section, base-stock level, zi, is suppliers’ decisions in our experiments. If i denotes the set of supplier i’s possible decisions, we can then express the choice probabilities using a logit form as follows:

$$\Pr (z_{i} ) = \frac{{\exp \left[ {\frac{{EU_{i} (z_{i} )}}{\beta }} \right]}}{{\sum\nolimits_{{\omega_{i} \in \Omega_{i} }} {\exp \left[ {\frac{{EU_{i} (\omega_{i} )}}{\beta }} \right]} }}\,,\,\,\,\,\,\,\forall z_{i} \in \Omega_{i} .$$

EUi (zi) in Eq. (1) denotes the expected utility of supplier i resulted from choosing a base-stock level zi over all possible decisions of the competitor. Parameter β, the bounded rationality parameter, shows supplier’s cognitive limitation in choosing the best decision. It is obvious when \(\beta \to \infty\), decisions follow a uniform distribution, and all decisions have an equal probability of occurrence. Conversely, β = 0 shows perfect rationality when a supplier always chooses the best decision with a probability of one.

In repetitive games, subjects become more experienced as the game proceeds (Bolton & Katok, 2008; Bostian et al., 2008), so they can make better decisions (Chen et al., 2012). To model this learning behavior and factor (b), we assume random error in subjects’ decisions decreases over the rounds. Hence, β is assumed to be a function of decision round, as shown in the following equation:

$$\beta (t) = \gamma + (\alpha - \gamma ){\text{e}}^{{ - \theta \left( {t - 1} \right)}}$$

in which α is the initial value of β, γ is the eventual value of β, θ is the rate of learning, and t is the decision round of the experiment.

In addition to making random errors in reaching their targeted decisions, the supplier might target decisions that are not meant to maximize their expected profit. In other words, suppliers might not be pure profit maximizers, rather they might include other behavioral factors in their utility function. One such behavior that has been studied in the literature in competitive setups is other-regarding and it is factor (c) here. Other-regarding assumes decision makers’ decisions are affected by their competitors’ observed payoffs. Fehr and Schmidt (1999) and Bolton and Ockenfels (2000) have studied other-regarding as a form of social preference and proposed models in which players’ utility includes relative payoff terms. In these models, players evaluate their own performance compared to the payoffs of other players and show either cooperative, competitive, or fairness preferences. Following these studies, we define supplier i's utility as

$$U_{i} (z_{i} ,z_{j} ) = \pi_{i} (z_{i} ,z_{j} ) + \sigma \left[ {\pi_{i} (z_{i} ,z_{j} ) - \pi_{j} (z_{i} ,z_{j} )} \right]^{ + } + \eta \left[ {\pi_{j} (z_{i} ,z_{j} ) - \pi_{i} (z_{i} ,z_{j} )} \right]^{ + }$$

in which, \(x^{ + } = \max (x,0)\), and σ and η capture supplier i's preference on the relative payoff when it outperforms the competitor j and when the competitor has a higher payoff, respectively. For values of σ < 0 and η < 0, subjects are fair-minded. If σ > 0 and η < 0, subjects have competitive preferences. Cooperation happens when −1 < σ < 0 and 0 < η < 1.

To find the values of the parameters in each model that creates the best fit with the experimental data, we use the Maximum Likelihood Estimation (MLE) method. Let the total number of subjects be m and each subject make a total of T decisions (T = 30 in our experiments). We can then define \({\mathbf{\mathbb{Z}}} = \left\{ {z_{i}^{t} |i = 1,...,m;t = 1,...,T} \right\}\) as the set of all base-stock decisions that subjects make during an experiment. The logarithmic form of the likelihood function can then be written as:

$$L\left( {\alpha ,\gamma ,\theta ,\sigma ,\eta ,\varepsilon |{\mathbf{\mathbb{Z}}}} \right) = \sum\limits_{i = 1}^{m} {\sum\limits_{t = 1}^{T} {\ln \left[ {\left( {1 - \varepsilon } \right).\Pr \left( {z_{i}^{t} } \right) + \varepsilon .\frac{1}{\left| \Omega \right|}} \right]} } ,$$

where \(\Pr (z_{i}^{t} )\) is calculated by Eq. (1). In this model, the likelihood function has been modified by adopting a unified error term, ε, which shows the probability of choosing the base-stock level in a completely random fashion. This means that QRE model is followed with a probability of 1 − ε, while, with a probability of ε, subjects choose their base-stock levels from a uniformly distributed decision set. We can think of this term as the impact of all other factors not included in the utility function (Harless & Camerer, 1994).

Results of the Behavioral Models

Table 3 shows the results of our models for the first three behavioral factors described above. We analyze each level of competition intensity separately. ESM Appendix A6 presents the results of this model at a lower level and shows the behavioral model parameters at each competition intensity for identical and non-identical suppliers separately.

Table 3 Behavioral model parameter estimates

The results show that subjects’ rationality level and learning behavior differ between the three competition intensities, and more importantly, subjects have different other-regarding preferences. The coefficients of the relative payoffs demonstrate that competitive preference prevails among the subjects under Low- and Medium-Intensity competitions (σ > 0 and η < 0). Indeed, subjects in these two competitions are not fair-minded and prefer to gain more than their competitors while dislike to have a lower profit. This observation is consistent with findings of Hu et al. (2017) who study suppliers’ behavior in a capacity investment competition. In the High-Intensity competition, however, subjects’ preferences are different, and they show fairness concerns (σ < 0 and η < 0).

To explain this observation, it is helpful to examine the supplier’s theoretical profit functions. Figure 2 shows the profit functions under different competition intensities and cost structures. These figures show that subjects experience profit functions with different shapes in the three competition designs. Moreover, under High-Intensity competition, the profit could turn negative when decisions deviate from the theoretical best values. The latter feature of the High-Intensity competition suggests when the subjects are under the threat of losing money, they are more inclined toward a fair interaction.

Fig. 2

Theoretical profits of the suppliers. Identical suppliers: a low intensity, b medium intensity, c high intensity. Non-identical suppliers: d low intensity, e medium intensity, f high intensity

To examine the validity of this argument, we conducted an additional experiment similar to our High-Intensity competition experiment. In this new experiment, the suppliers are endowed with an additional 25 units of profit. This prevents the suppliers’ profit to turn into negative values for the majority of decisions made by the subjects. The percentage of decisions resulting in a negative profit in this new experiment is 5%, compared to 44% in the original experiment.Footnote 1 This led to High-Intensity competition results with σ = 1.0 (> 0) and η = −0.4 (< 0), supporting the idea of competitive behavior when suppliers gain positive profit and fairness concern behavior when they are threatened by losing money.

The higher tendency of subjects toward fairness in loss conditions is consistent with findings from studies of ultimatum games which show the proposed and accepted offers in loss-framed games are higher than gain-framed (standard format) games (Buchan et al., 2005; Leliveld et al., 2009; Zhou & Wu, 2011). These studies relate the increase in proposers’ offers in loss sharing games to their unwillingness to hurt others and explain responders’ behaviors through their higher levels of inequality aversion when sharing a loss.


In the other-regarding models (Bolton & Ockenfels, 2000; Fehr & Schmidt, 1999), the assumption is that players’ decisions are affected by their competitors’ profit (outcome). Here, we examine whether decisions are also affected by the competitors’ decisions (action).

Based on the results presented in Fig. 1 and Table1 for the non-identical suppliers, Table 4 compares the difference between the decisions of high-cost and low-cost subjects in our experiments with theoretical predictions.

Table 4 Decision difference between high- and low-cost suppliers

This difference in experimental results is smaller than (or at most equal to) the theoretical predictions. This observation suggests that non-identical subjects tend to change their decisions to be closer to their competitors’ decisions, which could be the consequence of subjects’ imitation tendency.

Studies of the imitation behavior in the psychology and neuroscience literature demonstrate that individuals tend to respond in the direction of observed actions performed by another individual (Akins et al., 2002; Kilner & Lemon, 2013; Rizzolatti & Craighero, 2004). Social Learning Theory, as one of the well-studied theories of learning process and social behavior, supports these findings and argues that new behaviors can be acquired by observing and imitating others (Bandura & Walters, 1977; Miller & Dollard, 1941).

We name the subjects’ tendency toward imitating their competitor’s decisions in our experiments rival-chasing. To investigate the rival-chasing behavior and to evaluate how subjects change their decisions with respect to their competitors’ decisions, we define the parameter \(\varphi_{i}^{t}\) as the product of the difference between subject i’s decision from the previous round \((z_{i}^{t} - z_{i}^{t - 1} )\) and the difference between subject’s and competitor’s decisions in the previous round \((z_{j}^{t - 1} - z_{i}^{t - 1} )\). That is,

$$\varphi_{i}^{t} = (z_{i}^{t} - z_{i}^{t - 1} )(z_{j}^{t - 1} - z_{i}^{t - 1} ).$$

A positive value of \(\varphi_{i}^{t}\) means subject i changes the decision in round t \((z_{i}^{t} )\) toward the competitor’s decision that was made in the previous round \((z_{j}^{t - 1} )\), which supports rival-chasing behavior. On the other hand, a negative value of \(\varphi_{i}^{t}\) means the subject changes the decision away from the competitors’ decision in the previous round, which is evidence against rival-chasing. When \(\varphi_{i}^{t} = 0\), subject i either repeats his decision from the previous round, or subject i and competitor j have made the same decision in round t − 1.

Table 5 compares the percentages of suppliers’ decisions moving toward and away from their competitors’ decisions (ignoring all cases with \(\varphi_{i}^{t} = 0\))\(.\) Our results show that the proportion of decisions moving toward the competitor’s decision is significantly greater than the decisions moving away from the competitor’s decisions in all types of competition, which demonstrates a clear rival-chasing behavior. In identical competitions, the two suppliers have the same equilibrium points. Hence, rival-chasing can lead to a faster convergence of decisions. In non-identical competitions, however, the two supplier types have different equilibrium points. Rival-chasing behavior, in these cases, can justify the closer gaps between observed decisions compared with the gaps between the equilibrium points predicted by the theory.

Table 5 Direction of decision changes with respect to competitor’s most recent decision

ESM Appendix A7 presents further analysis of the rival-chasing behavior considering the relative profit of the players, and ESM Appendix A8 presents a mixed model analysis of the experimental results.


This research contributes to the behavioral operations management literature by experimentally examining the impact of competition intensity on the behavior of the competing suppliers in an outsourcing problem. We also examine how suppliers’ cost heterogeneity affects their decisions.

Our results demonstrate that suppliers provide a higher service level for the buyer under higher competition intensity levels. Moreover, suppliers with a lower cost structure have advantage of offering a higher service level. However, our experimental results show some patterns in subjects’ decisions that cannot be explained by the theoretical predictions. Moreover, we find that subjects’ behaviors vary according to the competition intensity and their cost structure.

To explain the observed behaviors, we develop our behavioral models based on bounded rationality, learning, and other-regarding preferences. The results show that subjects make different levels of random errors in their decisions under different competition intensities. Competition intensity also affects subjects’ learning behavior. Considering subjects’ relative profit concern in their utility function shows that other-regarding behavior not only plays an important role in the subjects’ decisions but also differs under different competition intensities. Subjects, in our competition setups, show competitive behavior when they are making profit; however, when the competition could result in negative profits, subjects become fair-minded.

We also demonstrate that subjects’ decisions are affected not only by their competitors’ relative payoffs, but also by the competitors’ decisions. Subjects monitor the competitors’ actions and their consequences to adjust their own decisions. This adjustment leads to the rival-chasing behavior, which is more frequent when the competitor is more successful and earns higher profits.

Our study also has implications for practice. In particular, the findings can help buyers and supply chain managers better understand the underlying mechanisms in competitive decision-making processes, and in turn, help them formulate more effective approaches to sourcing.

This study points to several productive directions for future research. Among these are investigating behavioral influences with different competition setups and criteria, as well as work to analyze more potential interactions of the behavioral factors considered in this study. We believe that further research of the new behavioral factor presented in this study, rival-chasing, would be also interesting.

Key Questions Reflecting Applicability in Real Life

  1. 1.

    How can buyers design their outsourcing mechanism to maximize the service level they receive?

  2. 2.

    How are bidding decisions that suppliers make influenced by aspects of competition?

  3. 3.

    In a supply base with heterogeneous players, how the service levels provided by the more efficient suppliers are different from the less efficient suppliers?

  4. 4.

    How decisions of the competing suppliers can be modeled and incorporated into the decision support systems to provide the managers with more accurate recommendations?

Data availability

Not applicable.

Code availability

Not applicable.


  1. 1.

    This percentage under Low- and Medium-Intensity competitions is 0% and 5%, respectively.


  1. Akins, C. K., Klein, E. D., & Zentall, T. R. (2002). Imitative learning in Japanese quail (Coturnix japonica) using the bidirectional control procedure. Animal Learning & Behavior, 30(3), 275–281.

    Article  Google Scholar 

  2. Anderson, L. R., & Freeborn, B. A. (2010). Varying the intensity of competition in a multiple prize rent seeking experiment. Public Choice, 143, 237–254.

    Article  Google Scholar 

  3. Bandura, A., & Walters, R. H. (1977). Social Learning Theory. Prentice Hall.

    Google Scholar 

  4. Benjaafar, S., Elahi, E., & Donohue, K. L. (2007). Outsourcing via service competition. Management Science, 53(2), 241–259.

    Article  Google Scholar 

  5. Bolton, G. E., & Katok, E. (2008). Learning by doing in the newsvendor problem: A laboratory investigation of the role of experience and feedback. Manufacturing and Service Operations Management, 10(3), 519–538.

    Article  Google Scholar 

  6. Bolton, G. E., & Ockenfels, A. (2000). ERC: A theory of equity, reciprocity, and competition. American Economic Review, 90(1), 166–193.

    Article  Google Scholar 

  7. Bostian, A. J. A., Holt, C. A., & Smith, A. M. (2008). Newsvendor “pull-to-center” effect: Adaptive learning in a laboratory experiment. Manufacturing and Service Operations Management, 10(4), 590–608.

    Article  Google Scholar 

  8. Buchan, N., Croson, R., Johnson, E., & Wu, G. (2005). Gain and loss ultimatums. Advances in Applied Microeconomics, 13, 1–23.

    Article  Google Scholar 

  9. Cachon, G. P., & Zhang, F. (2007). Obtaining fast service in a queueing system via performance-based allocation of demand. Management Science, 53(3), 408–420.

    Article  Google Scholar 

  10. Cannella, S., Di Mauro, C., Dominguez, R., Ancarani, A., & Schupp, F. (2019). An exploratory study of risk aversion in supply chain dynamics via human experiment and agent-based simulation. International Journal of Production Research, 57(4), 985–999.

    Article  Google Scholar 

  11. Chen, J., Zhao, X., & Shen, Z.-J. (2015). Risk mitigation benefit from backup suppliers in the presence of the horizontal fairness concern. Decision Sciences, 46(4), 663–696.

    Article  Google Scholar 

  12. Chen, J., Zhou, Y. W., & Zhong, Y. (2017). A pricing/ordering model for a dyadic supply chain with buyback guarantee financing and fairness concerns. International Journal of Production Research, 55(18), 5287–5304.

    Article  Google Scholar 

  13. Chen, Y., Su, X., & Zhao, X. (2012). Modeling bounded rationality in capacity allocation games with the quantal response equilibrium. Management Science, 58(10), 1952–1962.

    Article  Google Scholar 

  14. Chen, Y., & Zhao, X. (2015). Decision bias in capacity allocation games with uncertain demand. Production and Operations Management, 24(4), 634–646.

    Article  Google Scholar 

  15. Cooper, D. J., & Kagel, J. H. (2016). Other-regarding preferences a selective survey of experimental results. The Handbook of Experimental Economics.

    Article  Google Scholar 

  16. Cui, T. H., & Zhang, Y. (2017). Cognitive hierarchy in capacity allocation games. Management Science, 64(3), 1250–1270.

    Article  Google Scholar 

  17. Dasci, A., & Guler, K. (2019). Dynamic strategic procurement from capacitated suppliers. Production and Operations Management, 28(4), 990–1009.

    Article  Google Scholar 

  18. Donohue, K., Katok, E., & Leider, S. (2018). The Handbook of Behavioral Operations. Wiley.

    Book  Google Scholar 

  19. Du, S., Nie, T., Chu, C., & Yu, Y. (2014). Newsvendor model for a dyadic supply chain with Nash bargaining fairness concerns. International Journal of Production Research, 52(17), 5070–5085.

    Article  Google Scholar 

  20. Dubois, A., & Fredriksson, P. (2008). Cooperating and competing in supply networks: Making sense of a triadic sourcing strategy. Journal of Purchasing and Supply Management, 14(3), 170–179.

    Article  Google Scholar 

  21. Elahi, E. (2013). Outsourcing through competition: What is the best competition parameter? International Journal of Production Economics, 144(1), 370–382.

    Article  Google Scholar 

  22. Fahimnia, B., Pournader, M., Siemsen, E., Bendoly, E., & Wang, C. (2019). Behavioral operations and supply chain management—A review and literature mapping. Decision Sciences, 50(6), 1127–1183.

    Article  Google Scholar 

  23. Fehr, E., & Schmidt, K. M. (1999). A theory of fairness, competition, and cooperation. Quarterly Journal of Economics, 114(3), 817–868.

    Article  Google Scholar 

  24. Feng, T., & Zhang, Y. (2017). Modeling strategic behavior in the competitive newsvendor problem: An experimental investigation. Production and Operations Management, 26(7), 1383–1398.

    Article  Google Scholar 

  25. Giunipero, L. C., Bittner, S., Shanks, I., & Cho, M. H. (2019). Analyzing the sourcing literature: Over two decades of research. Journal of Purchasing and Supply Management, 25(5), 100521.

    Article  Google Scholar 

  26. Gundlach, G. T., Frankel, R., & Krotz, R. T. (2019). Competition policy and antitrust law: Implications of developments in supply chain management. Journal of Supply Chain Management, 55(2), 47–67.

    Article  Google Scholar 

  27. Harless, D.W., Camerer, C.F. (1994). The predictive utility of generalized expected utility theories. Econometrica, 62(6), 1251–1289.

    Article  Google Scholar 

  28. Heese, H. S., Kemahlıoğlu-Ziya, E., & Perdikaki, O. (2020). Outsourcing under competition and scale economies: When to choose a competitor as a supplier. Decision Sciences.

    Article  Google Scholar 

  29. Hu, S., Wan, Z., Ye, Q., & Chi, W. (2017). Supplier behavior in capacity investment competition: An experimental study. Production and Operations Management, 26(2), 273–291.

    Article  Google Scholar 

  30. Jain, T., & Hazra, J. (2017). Dual sourcing under suppliers’ capacity investments. International Journal of Production Economics, 183, 103–115.

    Article  Google Scholar 

  31. Jain, T., Hazra, J., & Cheng, T. C. E. (2020). Bidding for outsourcing contracts with capacity investments and cost asymmetry. Journal of the Operational Research Society, 71(12), 1986–2012.

    Article  Google Scholar 

  32. Jin, Y., Hu, Q., Kim, S. W., & Zhou, S. X. (2019). Supplier development and integration in competitive supply chains. Production and Operations Management, 28(5), 1256–1271.

    Article  Google Scholar 

  33. Johnsen, L. C., Voigt, G., & Corbett, C. J. (2019). Behavioral contract design under asymmetric forecast information. Decision Sciences, 50(4), 786–815.

    Article  Google Scholar 

  34. Kalkanci, B., Chen, K. Y., & Erhun, F. (2011). Contract complexity and performance under asymmetric demand information: An experimental evaluation. Management Science, 57(4), 689–704.

    Article  Google Scholar 

  35. Katok, E., & Pavlov, V. (2013). Fairness in supply chain contracts: A laboratory study. Journal of Operations Management, 31(3), 129–137.

    Article  Google Scholar 

  36. Kilner, J. M., & Lemon, R. N. (2013). What we know currently about mirror neurons. Current Biology, 23(23), R1057–R1062.

    Article  Google Scholar 

  37. Kumar, S. (2021). Structural changes and competitiveness of the steel industry in India: Assessed by productivity growth post economic reforms. International Journal of Global Business and Competitiveness, 16(1), 39–50.

    Article  Google Scholar 

  38. Leliveld, M. C., van Beest, I., van Dijk, E., & Tenbrunsel, A. E. (2009). Understanding the influence of outcome valence in bargaining: A study on fairness accessibility, norms, and behavior. Journal of Experimental Social Psychology, 45(3), 505–514.

    Article  Google Scholar 

  39. Li, C. (2013). Sourcing for supplier effort and competition: Design of the supply base and pricing mechanism. Management Science, 59(6), 1389–1406.

    Article  Google Scholar 

  40. Li, C., & Wan, Z. (2017). Supplier competition and cost improvement. Management Science, 63(8), 2460–2477.

    Article  Google Scholar 

  41. Li, W., Chen, J., & Chen, B. (2020). Sourcing strategy of original equipment manufacturer with quality competition. Decision Sciences, 51(5), 1110–1130.

    Article  Google Scholar 

  42. Li, Z., Ryan, J. K., Shao, L., & Sun, D. (2015). Supply contract design for competing heterogeneous suppliers under asymmetric information. Production and Operations Management, 24(5), 791–807.

    Article  Google Scholar 

  43. Liker, J. K., & Choi, T. Y. (2004). Building deep supplier relationships. Harvard Business Review, 82(12), 104–113.

    Google Scholar 

  44. McKelvey, R. D., & Palfrey, T. R. (1995). Quantal response equilibria for normal form games. Games and Economic Behavior, 10(1), 6–38.

    Article  Google Scholar 

  45. Miller, N. E., & Dollard, J. (1941). Social learning and imitation. Yale University Press.

  46. Momaya, K. S. (2020). The past and the future of competitiveness research: a review in an emerging context of innovation and EMNEs. International Journal of Global Business and Competitiveness, 14(1), 1–10.

    Article  Google Scholar 

  47. Ovchinnikov, A., Moritz, B., & Quiroga, B. F. (2015). How to compete against a behavioral newsvendor. Production and Operations Management, 24(11), 1783–1793.

    Article  Google Scholar 

  48. Ren, H., & Huang, T. (2018). Modeling customer bounded rationality in operations management: A review and research opportunities. Computers & Operations Research, 91, 48–58.

    Article  Google Scholar 

  49. Rizzolatti, G., & Craighero, L. (2004). The mirror-neuron system. Annual Review of Neuroscience, 27, 169–192.

    Article  Google Scholar 

  50. Shee, H. K., VanGramberg, B., & Foley, P. (2011). Capability and practices to enhance firm competitiveness under competitive intensity: A SEM approach. International Journal of Global Business and Competitiveness, 6(1), 1–13.

    Google Scholar 

  51. Shinde, A., Mehta, P., & Amit, R. K. (2020). An experimental investigation of newsvendor decisions under ambiguity. International Journal of Production Research.

    Article  Google Scholar 

  52. Sonar, H. C., Khanzode, V., & Akarte, M. (2020). A conceptual framework on implementing additive manufacturing technology towards firm competitiveness. International Journal of Global Business and Competitiveness, 15(2), 121–135.

    Article  Google Scholar 

  53. Song, J., Zhang, J., & Cheng, T. C. E. (2020). Inventory management and the value of quick response to the retailer facing boundedly rational strategic customers. International Journal of Production Research.

    Article  Google Scholar 

  54. Song, Y., & Zhao, X. (2017). A newsvendor problem with boundedly rational strategic customers. International Journal of Production Research, 55(1), 228–243.

    Article  Google Scholar 

  55. Villa, S., & Castañeda, J. A. (2018). Transshipments in supply chains: A behavioral investigation. European Journal of Operational Research, 269(2), 715–729.

    Article  Google Scholar 

  56. Wetzstein, A., Feisel, E., Hartmann, E., & Benton, W. C. (2019). Uncovering the supplier selection knowledge structure: A systematic citation network analysis from 1991 to 2017. Journal of Purchasing and Supply Management, 25(4), 100519.

    Article  Google Scholar 

  57. Xu, X., Ji, Y., Bian, Y., & Sun, Y. (2016). Service outsourcing under co-opetition and information asymmetry. Journal of the Operational Research Society, 68(6), 666–677.

    Article  Google Scholar 

  58. Zhao, X., Xue, L., & Zhang, F. (2014). Outsourcing competition and information sharing with asymmetrically informed suppliers. Production and Operations Management, 23(10), 1706–1718.

    Article  Google Scholar 

  59. Zhao, Y., & Zhao, X. (2016). How a competing environment influences newsvendor ordering decisions. International Journal of Production Research, 54(1), 204–214.

    Article  Google Scholar 

  60. Zhao, Y., & Zhao, X. (2017). On elicitation-method effect in game experiments: a competing newsvendor perspective. Journal of the Operational Research Society, 69(4), 1–17.

    Article  Google Scholar 

  61. Zhou, X., & Wu, Y. (2011). Sharing losses and sharing gains: Increased demand for fairness under adversity. Journal of Experimental Social Psychology, 47(3), 582–588.

    Article  Google Scholar 

Download references


We would like to thank the anonymous reviewers and editors for their feedback. Their comments and suggestion helped us to enhance the strength of our research. We would like to also thank the University of Massachusetts Boston who participated in our experiments and provided the data on which our paper is based.


No funding was received for conducting this study.

Author information




All authors conducted experiments, participated in the statistical analysis, wrote sections of the paper, and worked on editing it. Each of the authors has approved the manuscript in the critical appraisal and revisions of it.

Corresponding author

Correspondence to Mohsen Ahmadian.

Ethics declarations

Conflict of interest

The authors declare that there is no conflict of interest in this research.

Ethics approval

The authors confirm that the study was granted exemption by the University of Massachusetts Boston IRB and certify that the study was performed in accordance with the ethical standards as laid down in the 1964 Declaration of Helsinki and its later amendments or comparable ethical standards.

Supplementary Information

Below is the link to the electronic supplementary material.

Supplementary file1 (PDF 727 kb)

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Ahmadian, M., Elahi, E. & Blake, R. The Behavioral Effects of Competition Intensity and Cost Structure on Competing Suppliers: An Experimental Study in the Context of the USA. JGBC (2021).

Download citation


  • Supply competition
  • Behavioral operations management
  • Experimental economics
  • Outsourcing
  • Competition intensity