Impact of Social Preferences,Trust and Behavioural Norms on Cooperation: Experimental Evidence from Prisoner’s Dilemma Game

Abstract

The main objective of the study is to evaluate the impact of social preferences, trust and behavioural norms on strategic choices made by economic agents using the framework of a one-shot Prisoner’s Dilemma (PD) game. The standard neoclassical economics model predicts that all players will choose a strategy of non-cooperation, establishing it as the Nash equilibrium. We examine how social preferences and behavioural norms influence the player’s expectations from others and their own choices, by experimentally testing for the same using a 2 × 2 factorial design with the following three levels of partner selection arrangements: (i) game playing with stranger, (ii) face-to-face game playing with peer, and (iii) game playing with friend. The experimental results reveal significant deviations from the standard predictions of the one-shot PD game. The experimental data revealed that (i) players while interacting with a peer cooperated significantly more than playing with a stranger and (ii) players playing with a friend cooperated significantly more than playing with a peer. These changes in the social preferences of the players were further investigated in terms of trust/distrust and four behavioural norms—positive reciprocity, negative reciprocity, exploitation and altruism, and how these factors interact in influencing decision-making with respect to cooperation.

JEL Classifications: C72, C91, D01, D91

Keywords

Behavioural norms cooperation Nash equilibrium Prisoner’s dilemma game social preferences trust

Introduction

Game theory provides a rich platform for developing and experimentally testing agent decision-making models. The standard neoclassical game theory models are based on the underlying assumption that economic agents are purely self-interested and are concerned with maximizing their own individual material payoffs. However, the assumption of purely self-interested behaviour is often contradicted in experimental findings and has led to the development of social preference models (Camerer et al., 2004; Charness & Rabin, 2002; Fehr & Fischbacher, 2002; Jordan et al., 2014). Results from the discipline of behavioural economics show that individuals tend to be influenced by social preferences and do not always maximize their individual material payoffs.

Within behavioural economics, different models of social preference have been put forth to account for the observed deviations from the standard predictions of game theory. Social preference theories typically assume that economic agents are rational, but the utility functions of individuals include more than just their own material payoff, that is, individuals are concerned for not only their own individual material payoffs but also the payoffs received by the other individuals. Two main types of social preference models have been proposed in the literature: altruistic preference models and reciprocal preference models. The altruistic preference models (Anderson et al., 1998; Andreoni, 1989; Andreoni & Miller, 1993; Levine, 1998) assume that individuals are willing to unconditionally cooperate and be kind towards others irrespective of their expectations about the other’s behaviour. The reciprocal preference models propose that individuals are kind to those who treat them kindly and are unwilling to be kind towards those who treat them unkindly (Bolton & Ockenfels, 2000; Fehr & Schmidt, 1999; Hayashi et al., 1999; Kiyonari et al., 2000; Rabin, 1993).

In recent years, researchers have provided experimental evidence that has questioned the ‘pure self-interest’ hypothesis assumed in the standard game-theoretic models by demonstrating that many people are motivated by concerns beyond their own individual material benefits. To study the logic of social dilemmas and understand real-world complexities, researchers have invented a variety of games. One of the most commonly studied social-dilemma games is the two-person prisoner’s dilemma (PD) game.

Experimental studies have provided evidence that in the one-shot PD game, significant proportion of decision-makers choose to cooperate under different social contexts: positive interactions between player’s (Rand et al., 2009); more information about the partner’s strategic choices (Martin et al., 2014); impact of trust and benevolence (Capraro et al., 2014) and preferential interactions with known partners (Hruschka & Henrich, 2006).

Cooperation in Prisoner’s Dilemma Game

In the literature, different theories of cooperation have been proposed (Fischbacher et al., 2001; Hammerstein, 2003), and various factors favourable or unfavourable to cooperation have been studied in laboratory experiments (Andreoni & Miller, 1993; Andreoni & Varian, 1999; Bereby-Meyer & Roth, 2006; Grimm & Mengel, 2009). Despite this long history of research using the PD, there is still no agreement on why people cooperate. This section highlights the various factors that affect the degree of cooperation in social dilemma games.

Reciprocity

Reciprocity is characterized as being ‘kind’ towards someone who is expected to be kind in return (positive reciprocity) and being ‘unkind’ towards someone who is expected to be unkind (negative reciprocity). This motivation to reciprocally act in a social dilemma game gives rise to two possible outcomes: One, where both individuals are kind towards one another, or the other, where both choose to be unkind. Any non-cooperative action of the opponent that provides unfair distribution of resources is interpreted as ‘unkind’, and the other player reciprocates this action with a defection strategy for being unkind (Greig & Bohnet, 2008). Thus, reciprocity is the propensity of the player to penalize the other player by decreasing his payoff when he is perceived to be ‘unkind’ and act generously by increasing the opponent’s payoff when he is believed to be ‘kind’ (Bolton & Ockenfels, 2000; Fehr & Schmidt, 1999). Recent experimental studies found that economic agents are more sensitive towards norms of negative reciprocity than positive reciprocity; it is considered relatively fair to punish defectors for not cooperating than rewarding the co-operators for their compliance (Chernyak et al., 2019; Shaw et al., 2019).

In the PD experimental games, the impact of norms of reciprocity was quite evident, wherein it was found that individuals were willing to cooperate to match their opponent’s good faith (Morris et al., 1998). Experimental evidence in the literature (Hayashi et al., 1999; Kiyonari et al., 2000) suggests that players, even in simultaneous one-shot PD games, behave reciprocally and expect the same to be reciprocated by their partners. It was observed that players yielded a more satisfactory outcome when they voluntarily chose to cooperate than defect when their opponent cooperated. Rabin (1993) makes a similar prediction on the norms of reciprocity based on a ‘kindness’ function. It was observed that players are willing to sacrifice more money to be kind if they believe their opponent is also kind in return. Therefore, cooperation can be sustained as an equilibrium in the one-shot game if players trust their counterparts and are reciprocally motivated to cooperate.

Trust

Trust can be defined as the positive beliefs that one agent has about the intention of the other agents (Cox et al., 2001). Trusting behaviour is considered a vital factor that can reinforce cooperation in a one-shot PD game (Janssen, 2008; Ostrom & Walker, 2003). It has been found that cooperation can increase between anonymous agents in non-repeated games when the agents have the ability to recognize the trustworthiness of other agents. The greater the trust, the greater the positive externalities delivered through cooperation, which increases the welfare of the entire group. However, at the same time, the non-compliance and disloyalty by the other player will generate a greater degree of negative externalities through non-cooperation. The trust built through elicited promises between the agents was also found to be effective in nurturing the partnership agreements between the players (Chen & Zhang, 2021; Suneja & Das, 2022). According to Cook and Cooper (2003), players are initially reluctant co-operators, as they generally hesitate to cooperate due to the fear of non-compliance by the other players. The benefit of mutual cooperation can be materialized only if there is substantial assurance that the other group players will cooperate.

Altruism

Unlike the norms of reciprocity, altruism reflects unconditional cooperation and kindness towards the other players, irrespective of their intentions and choices (Andreoni, 1989). The key aspect that differentiates altruism from reciprocally motivated individuals is that an altruistically motivated person will never penalize the opponent for their strategic choices by decreasing their payoff. Rather, for an altruistic person, the utility increases with the increase in the well-being of the other players (Baumert et al., 2014; Gintis et al., 2003). The evolution of the norms of altruism is considered a significant factor for individuals’ intellectual and moral development that can ensure a sustainable future (Mangone, 2020).

The study by Shafir and Tversky (1992) simulated the PD game on 80 undergraduate students and found that the degree of cooperation was 3% despite knowing the other person had defected in the game. Batson and Moran (1999) examined similar effects of pro-social behaviour and concluded that empathic emotion could arouse the altruistic motivation of the players to increase the other’s welfare. The study by Weiss and Peres (2014) also supported this clause and found that in the standard PD game, the egoist players who only valued their payoffs were at a disadvantageous position relative to the players who chose to cooperate for a mutually-advantageous payoff. It is believed that altruists tend to do better in strategic situations than those motivated solely by a concern for their well-being.

Using the framework of one-shot PD, we aim to investigate the influence of social preferences, trust and behavioural norms in influencing player’s strategic decisions using a single-shot simultaneous PD game.

Structure of the Simultaneous Single-Shot Two-Person Prisoner’s Dilemma (PD) Game

The overall strategy in the two-player one-shot simultaneous PD game as seen in Table 1 is based on two strategic choices—either to cooperate or to non-cooperate, based on the expected strategy by the other player.

Table 1.

Structure of the Game.

Prisoner’s	Dilemma Game	Player 2
		Cooperate (C)	Non-cooperate (NC)
Player 1	Cooperate (C)	Gain (G), Gain (G)	Sucker (S), Temptation (T)
	Non-cooperate (NC)	Temptation (T), Sucker (S)	Loss (L), Loss (L)

When both players opt for cooperation strategies (opt for C–C): then both individuals are rewarded for this cooperation. Each player receives a payoff denoted as ‘G’.

When both players do not cooperate, that is, they choose NC–NC, then both the individuals lose and this loss is denoted as payoff ‘L’: this is less than the payoff ‘G’, they could have earned if both of them had used the cooperation strategy.

If one player cooperates and the other does not, then the co-operator receives the lowest payoff ‘S’ (this is the Sucker payoff) and the defector receives the largest payoff ‘T’ (the Temptation payoff).

Payoff structure of the PD game: T>G>L>S

According to the standard game-theoretic assumptions used in the PD game, a rational individual is driven by the motive of individual material self-interest. Such a player is expected to non-cooperate if he expects the other player to cooperate, since by doing so he receives the Temptation payoff (T) which is greater than the Gain payoff (G). If he expects the other player to non-cooperate, he will choose to non-cooperate since by doing so, he receives the Loss payoff (L) which is greater than the Sucker payoff (S). The other player in the game is assumed to play in a similar way. Hence, both players are expected to non-cooperate and this establishes mutual non-cooperation (NC–NC) as the Nash equilibrium strategy. This equilibrium yields a Pareto-inferior outcome for both the players: both players receive a payoff that is less than what they would have received had they both cooperated.

In our research, we shed light on the strategic motivations of the individuals to cooperate by experimentally exploring the one-shot PD game. The strategic interaction between the players involves forming expectations about other player’s beliefs and preferences, and then reacting based on one’s own preferences. In order to make predictions about the outcome of the PD game, the interplay between these two factors has to be taken into account. Our study focuses on how different partner selection agreements embodying varying degrees of social affinity change player’s expectations and social preferences, which determine their strategic choices in the simultaneous one-shot PD game.

Methodology

Design of the Experiment

The experimental design is based on a two-player simultaneous one-shot PD game that evaluates the degree of cooperation among players. In our experiment, the players were made to play a one-shot symmetric game with the payoff structure as provided in Table 2. To avoid framing effect on the player’s strategic choices, the strategy to cooperate by the two individuals is coded as A₁ and B₁, respectively, and the non-cooperation strategy as A₂ and B₂. Given this payoff matrix, the option to mutually cooperate, that is, opting for strategy A₁―B₁ by both the players yields a payoff of 100 units to both players, whereas the mutual non-cooperation strategy of A₂―B₂ yields a payoff of 50 each. In case one player chooses to non-cooperate while the other player chooses to cooperate (A₂―B₁ or A₁―B₂), then the player who cooperates gets a sum of 30 units, while the player who non-cooperates gets 120 units.

Table 2.

Payoff Matrix of Prisoner Dilemma Game.

		Other Player’s Strategy
Your strategy		B₁	B₂
	A₁	You: 100Other player: 100	You: 30Other player: 120
	A₂	You: 120Other player: 30	You:50Other player: 50

Traditionally in experimental economics, material incentives are commonly used to motivate decisions. However, in some cases, monetary incentives are found to be ineffective and observed to have no significant difference in the strategic behaviour of the economic agents. The study by Camerer and Hogarth (1999) found that the effects of incentives led to a ‘mixed and complicated’ framework, and the outcome ‘depended on the kind of task a person performs’. The pioneers of experimental economics, Kahneman and Tversky (1979), also stressed that an incentive-compatible payment schedule is not required to elicit a true response from the economic agents and stated that ‘people often know how they would behave in actual situations of choice, and the subjects have no special reason to disguise their true preferences’. Researchers have justified that the participants can make social judgements and are able to accurately report their true beliefs about the other player and determine their strategic choices in the social dilemma games even by using hypothetical incentives.

The study by Read (2005) highlighted that monetary incentives influence the strategic behaviour of the economic agents through—(i) Cognitive Exertion: the increase in the cognitive load due to the amount of thought put into the task that determines the final strategic outcome of the game. The complex payoff structure can become cognitively taxing for the participants to comprehend the game while making strategic decisions (Charness et al., 2021); (ii) Motivational Focus: the incentive changes the agent’s goals. In a PD game, the incentivized structure guides the players to recognize the payoff dominance structure of the game and disillusions them to believe that defection is the ideal strategy. The participants may have a strong intrinsic motivation to cooperate, but when they are insufficiently compensated for doing so, they end up with non-cooperation. Hence, by merely paying people, we are encouraging them to think harder about the incentive structure and directing their decisions towards our predetermined experimental objective; (iii) Emotional Triggers: the agent’s instantaneous response for a real and hypothetical payoff condition. The study by Cummings et al. (1995) found that the participant’s emotional arousal in response to potential loss or gain of money through hypothetical choices were quite evident. Researchers were able to elicit an intuitive response from the player, even when the players were unable to experience or predict their responses to the real incentives.

Since monetary incentives are neither necessary nor sufficient for achieving the research objective, we can conclude that the experimental results can be derived from observed behaviour by eliciting perceived social preferences and norms directly from subjects. Using a non-incentivized framework for our experiment, we capture the participant’s perception of the other player’s strategic response and their choice to cooperate or non-cooperate in the game. In our experiment, we further investigate whether the participant’s beliefs and strategic choices are influenced by social affinity by employing different partner treatment conditions. Using the proposed experimental framework without monetary incentives, the player’s expectation about the other player’s strategy and their strategy in response is ascertained by asking the following question to all the participants in the survey:

(A) Expected strategy of the other player

We evaluate the player’s perception of the expected strategy of the other player, by asking all respondents to answer the following question:

How do you expect the other player to play?

Cooperate or Non-cooperate

(B) Strategy in response to the other player

With their perception about how they expect the other player to play, we investigate whether the player will cooperate or non-cooperate by asking the following:

Your strategy?

Cooperate or Non-cooperate

Experimental Treatments

In our research, we investigate the player’s expectations and their own decisions under different experimental treatments with the following three levels of partner selection arrangements: (i) game playing with stranger, (ii) face-to-face game playing with peer and (iii) game playing with friend.

Treatment 1: Stranger

In this case, the participant plays with a stranger. With the assumption of negligible emotional bonding and social affinity between the players under this treatment condition, the respondent is asked about how he expects the other player to play and about his own choice between cooperating and non-cooperating:

How do you expect the other player to play?

Cooperate or Non-cooperate

Your strategy? Cooperate or Non-cooperate

Treatment 2: Peer Interaction

In this case of peer interaction condition, the participants were asked to play the game with the person sitting behind them in the classroom. The players were forbidden from communicating with each other. The participants were asked the following:

You will now play the game with the person sitting behind you.

What do you expect your partner to do? Cooperate or Non-cooperate

How will you play? Cooperate or Non-cooperate

Treatment 3: Friend

This treatment assumes the maximum degree of social affinity, wherein the participants were required to play with a friend. The respondents were asked the following:

Assume that you are playing with your friend.

What do you expect your friend to do? Cooperate or Non-cooperate

How will you play? Cooperate or Non-cooperate

Hypotheses

Based on these treatment conditions with respect to the three partner selection arrangements as above, we propose the following hypotheses that are then empirically tested:

H₁: The expected cooperation from the other player increases with social affinity.

H_1a: The expectation of cooperation from the other player is greater when playing face-to-face with a peer relative to playing with a stranger.

H_1b: The expectation of cooperation from the other player is greater in playing with a friend relative to playing face-to-face with a peer.

H₂: The frequency of playing the cooperation strategy increases with social affinity.

H_2a: The frequency of playing the cooperation strategy is greater when playing face-to-face with a peer relative to playing with a stranger.

H_2b: The frequency of playing the cooperation strategy is greater when playing with a friend relative to playing face-to-face with a peer.

Sample

The PD experiment was conducted on a pool of 436 undergraduate and postgraduate students of different educational backgrounds (engineering, arts, science, business, social sciences and medicine from premier Indian universities) based in New Delhi. The sample involves a mixed pool of students currently pursuing graduate or postgraduate courses (nearly 71% with work experience of 0–10 years) and executives pursuing career advancement courses (nearly 29% with work experience of 10 to more than 20 years). The demographic profile of the sample is such that the average age of participants is 30 years, with nearly 63% male and 37% female participants. Table 3 summarizes the profile of the participants.

Table 3.

Sample Characteristics.

Respondent Characteristics	No. of Participants	Percentage of Participants
Gender
Male	274	63%
Female	162	37%
Educational background
Engineering	192	44%
Arts	65	15%
Science	58	13%
Business/commerce	51	12%
Social Science	35	8%
Medicine	25	6%
Others	10	2%
Years of work experience
0–5	124	28%
5–10	190	44%
10–15	64	15%
15–20	46	11%
More than 20	12	3%
Average age of participants = 30

Results

The results from our experimental games are summarized as follows.

Stranger versus Peer Treatment

(A) Expectation from the other player

From Table 4, we can observe that the frequency of expectation of cooperation from the other player increases from 217 (50%) when playing with a stranger to 261 (60%) while playing with a peer. Thus, playing in a face-to-face situation with one’s peer increases the frequency of expected cooperation from the other player by 10% points as compared to when one is playing with a stranger.

Table 4.

Expected Cooperation from the Stranger vs Peer.

	Expectation from the Other Player
Strategy	Peer	Stranger
Non-cooperate	175(40%)	219(50%)
Cooperate	261(60%)	217(50%)

Chi-Square Tests	Value	df	Asymp. Sig. (2-sided)	Exact Sig. (2-sided)	Exact Sig. (1-sided)
Pearson chi-square	8.964	1	.003
Continuity correction	8.561	1	.003
Likelihood ratio	8.980	1	.003
Fisher’s exact test				.003	.002

The difference in the frequency of expectation in cooperation from the other player in the peer viz-a-viz the stranger treatment is tested for statistical significance using the chi-square test for difference in proportions. The test confirms that the difference in the frequency of expected cooperation is statistically significant (Table 4). Our results hence confirm the hypothesis that the frequency of expected cooperation from the other player is greater for game playing with a peer relative to playing with a stranger (H_1a).

(B) Strategy in response to the other player

The frequency of playing the cooperation strategy increases from 122 (28%) while playing with a stranger to 230 (53%) when playing face-to-face with a peer. Using the chi-square difference in proportions test, we find that the difference in frequency of playing the cooperation strategy is statistically significant (Table 5). Our results hence confirm the hypothesis that the frequency of playing the cooperation strategy is greater in the case of playing face to face with a peer relative to playing with a stranger (H_1b).

Table 5.

Frequency of Cooperation with Stranger vs Peer.

Response to the Other Player
Strategy	Peer	Stranger
Non-cooperate	206(47%)	314(72%)
Cooperate	230(53%)	122(28%)

Chi-Square Tests	Value	df	Asymp. Sig. (2-sided)	Exact Sig. (2-sided)	Exact Sig. (1-sided)
Pearson chi-square	55.567	1	.000
Continuity correction	54.543	1	.000
Likelihood ratio	56.272	1	.000
Fisher’s exact test				.000	.000

Peer versus Friend Treatment

(A) Expectation from the other player

Playing with a friend compared to playing face to face with a peer can be assumed to raise the degree of social affinity and emotional connection, and hence raise the frequency of cooperation expected from the other player (H_2a). From Table 6, it can be seen that the expectation from the other player to cooperate increases from 261 (60%) when playing face to face with a peer to 345 (79%) while playing with a friend. The chi-square test for the difference in proportions establishes that the difference is statistically significant. Our results hence support H_2a.

Table 6.

Expected Cooperation from the Peer vs Friend.

Expectation from the Other Player
Strategy	Friend	Peer
Non-cooperate	91(21%)	175(40%)
Cooperate	345(79%)	261(60%)

Chi-Square Tests	Value	df	Asymp. Sig. (2-sided)	Exact Sig. (2-sided)	Exact Sig. (1-sided)
Pearson chi-square	38.170	1	.000
Continuity correction	37.266	1	.000
Likelihood ratio	38.667	1	.000
Fisher’s exact test				.000	.000

(B) Strategy in Response to the other player

Owing to greater social affinity, we expect the frequency of playing the cooperation strategy to be higher when playing with a friend than when playing face to face with a peer (H_2b). The statistically significant impact of increased social affinity on frequency of cooperation is reflected in Table 7, wherein we observe that the frequency of cooperation increases from 230 (53%) when playing face-to-face with a peer to 339 (78%) when playing with a friend. This confirms H_2b.

Table 7.

Frequency of Cooperation with Peer vs Friend.

	Response to the Other Player
Strategy	Friend	Peer
Non-cooperate	97(22%)	206(47%)
Cooperate	339(78%)	230(53%)

Chi-Square Tests	Value	df	Asymp. Sig. (2-sided)	Exact Sig. (2-sided)	Exact Sig. (1-sided)
Pearson chi-square	60.092	1	.000
Continuity correction	58.994	1	.000
Likelihood ratio	61.114	1	.000
Fisher’s exact test				.000	.000

Discussion

We observe that with increasing degree of social affinity among the players under the different partner treatments, the expectation of cooperation from the other player as well as the playing of the cooperation strategy by the player’s increases. Figure 1 below summarizes the strategic interaction of the players in the one-shot simultaneous PD game.

The frequency of cooperation expected from the other player is highest at 79% in case of playing with friend, and decreases to 60% and 50% while playing face to face with a peer and stranger, respectively. Increased social affinity thus proves to be effective in raising the level of trust among the players. Increased social affinity also raises the frequency of playing the cooperation strategy from 28% when playing with a stranger to 53% and 78% when playing with a peer and a friend, respectively.

Figure 1.

Cooperation by the Players in Prisoner’s Dilemma Game.

Evaluation of Trust and Behavioural Norms

To analyse the data further, we use a model of behavioural norms as shown in Figure 2 to understand the player’s motivation to cooperate in the game:

Figure 2.

Evaluation of Behavioural Norms in Prisoner’s Dilemma Game.

Based on the player’s preferences to cooperate or non-cooperate in the PD game in response to their expectation of how the other player will play (trusting/distrusting the other), the following four forms of behavioural norms emerge in our study: (i) positive reciprocity norm, which occurs when the player cooperates in response to the expectation that the other will cooperate, (ii) negative reciprocity norm, when the player non-cooperates in response to the expectation that the other player will non-cooperate, (iii) altruism, when the player cooperates in response to the expectation that the other player will non-cooperate and (iv) exploitation norm, when the player non-cooperates in response to the expectation that the other player will cooperate. This framework is further used to evaluate the impact of behavioural norms on the player’s strategic choice to either cooperate or non-cooperate, which is conditional on their expectation from the other player. We propose the following hypothesis which is empirically verified:

H₃: The behavioural norms influence the strategic choices of the players.

H_3a: The norms of negative reciprocity and exploitation negatively influence the strategic choice to cooperate.

H_3b: The norms of positive reciprocity positively influence the strategic choice to cooperate.

Table 8 summarizes the data with respect to trusting/distrusting the other and the four behavioural norms, in the context of the different partner selection treatments of stranger, peer and friend.

Table 8.

Trust-Distrust and Behavioural Norms with Respect to Different Partner Treatments.

Expectation from the Other Player	Your Response to the Other Player	Behavioural Norm		Stranger	Peer	Friend
Cooperate (C)	Cooperate (C)	Trust	Positive reciprocity	11225.68%	21749.77%	32574.54%
Cooperate (C)	Non-cooperate (NC)	Trust	Exploitation	10524.08%	4410.09%	204.58%
Non-cooperate (NC)	Non-cooperate (NC)	Distrust	Negative reciprocity	20947.93%	16237.15%	7717.66%
Non-cooperate (NC)	Cooperate (C)	Distrust	Altruism	102.29%	132.98%	143.21%

To test whether the differences observed in Table 8 in adoption of the four behavioural norms are statistically significant, the chi-square test for difference in proportions was employed. Our results support H₃ (Table 9), wherein it can be observed that the difference in proportions for the two groups (stranger→peer and peer→friend respectively) were statistically significant in the case of the norms of positive reciprocity, exploitation and negative reciprocity. While the preference to cooperate irrespective of the player’s belief that the other player may not cooperate (altruism) was found to be statistically insignificant for both groups (stranger–peer and peer–friend). We can hence conclude that social affinity has a statistically significant impact on behaviour of participants by influencing adoption of the behavioural norms of positive reciprocity, exploitation and negative reciprocity.

Table 9.

Significance of Behavioural Norms with Respect to Different Partner Treatments.

Chi-Square Tests: Positive Reciprocity
Stranger→Peer				Peer→Friend
	Value	df	P		Value	df	P
χ²	154	1	<.001	χ²	146	1	<.001
N	436			N	436
Chi-square tests: Exploitation
Stranger→Peer				Peer→Friend
	Value	df	P		Value	df	P
χ²	52.2	1	<.001	χ²	187	1	<.001
N	436			N	436
Chi-square tests: Negative reciprocity
Stranger→Peer				Peer→Friend
	Value	df	P		Value	df	P
χ²	280	1	<.001	χ²	158	1	<.001
N	436			N	436
Chi-square tests: Altruism
Stranger→Peer				Peer→Friend
	Value	df	P		Value	df	p
χ²	5.83	1	.063	χ²	6.46	1	.290
N	436			N	436

Deviation from Standard Prisoner’s Dilemma Game Predictions

As per the standard framework of the PD game, the strategic choice of the players to opt for a non-cooperation strategy is due to the implicit assumptions that all players will either play by the negative reciprocity norm or by the exploitation norm. In other words, if a player expects the other to non-cooperate, he will play non-cooperate, and if he expects the other to play cooperate, he will play non-cooperate, seeking to maximize his individual material payoff. Since the other player is also expected to play in similar fashion, standard game theory predicts the emergence of a Nash equilibrium of non-cooperate–non-cooperate. This prediction is expected to hold true no matter what the identity of the two players is since it is assumed that all players shall play to maximize their individual material payoff. It makes no difference to the expected outcome of the game whether the players are strangers, peers or friends. Even though the expected Nash-equilibrium outcome of non-cooperation–non-cooperation is Pareto-inferior to the cooperation–cooperation outcome, the standard assumptions of the PD game lead to the prediction that the non-cooperation–non-cooperation outcome will invariably prevail.

Contrary to the conventional predictions of the PD game, the results of our experiment suggest that besides the norms of negative reciprocity and exploitation, the norms of positive reciprocity and altruism were also prevalent in the game playing by the participants in our experiment. Figure 3 reflects the percentage frequency of various behavioural norms coupled with trust/distrust in our study, under different partner selection arrangements. It can be observed from Figure 3 that with increasing degree of social affinity among the players (stranger→peer→friend), the percentage of players exhibiting trust coupled with positive reciprocity increased from 26% with a stranger, to 50% with a peer and 75% with a friend. The frequency of distrust coupled with the norm of negative reciprocity declined to nearly 18% when playing with a friend, in comparison to 37% and 48% with a peer and a stranger, respectively. The frequency of trust coupled with the norm of exploitation diminished significantly with an increase in social affinity: only 5% of the participants were willing to exploit their friend in contrast to 10% and 24% of participants who were willing to exploit a peer and a stranger respectively. The frequency of distrust coupled with the choice of the norm of altruism marginally increased from 2% for strangers, 3% for peers and 3.2% for friends.

Figure 3.

Trust and Behavioural Norms with Partner Selection Treatments.

Conclusions

In our experimental setup, we highlight that the endogenous beliefs of the players cannot be generalized, and the individual differences cannot be captured through a single template of social preferences. The strategic interaction between the players have been evaluated by—First, assessing the players’ perception of the game, that is, their beliefs and expectation from the other players; and second, assessing how the players strategize and choose, given their perception of other’s strategy in the game. Using these two factors simultaneously, we identified the behavioural norms that influence player choice in a one-shot simultaneous PD game. The behavioural norms that emerged from the strategic interplay between the players based on their beliefs about the other players’ anticipated choice and their strategic choices in return—(i) positive reciprocity; (ii) exploitation; (iii) negative reciprocity; and (iv) altruism. It was found that the norms of positive reciprocity and altruism increased cooperation, while the norms of negative reciprocity and greed diminished cooperation. These norms coupled with trust/distrust influence the players decisions to cooperate or non-cooperate.

A key finding of our study is that people react differently under different treatment conditions that exemplify varying degrees of social affinity between the players. We investigated the impact of three partner selection arrangements on player’s strategic choices—(i) playing with stranger, (ii) playing face-to-face with peer and (iii) playing with friend. We discovered that both the expected cooperation from the other player and the player’s frequency of cooperation increased significantly with greater social affinity. In particular, the adoption of the norm of positive reciprocity increased significantly with increase in social affinity while the adoption of the norm of exploitation declined. We can hence conclude that social preferences have the potential to significantly influence norm-based decision-making of agents and encourage cooperative behaviour. This finding has wide-ranging and important policy implications in a variety of contexts (economic/social/political) wherein one seeks to analyse or encourage cooperative behaviour.

Footnotes

Consent to Participate

Verbal informed consent was obtained from all individual participants included in the study.

Data Availability

The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Declaration of Conflicting Interests

The authors declared no potential conflicts of interest with respect to the research, authorship and/or publication of this article.

Funding

The authors received no financial support for the research, authorship and/or publication of this article.

ORCID iDs

Debashree Das

References

Anderson

S. P.

, Goeree

J. K.

, & Holt

C. A.

(1998). A theoretical analysis of altruism and decision error in public goods games. Journal of Public Economics , 70(2), 297–323. https://doi.org/10.1016/S0047-2727(98)00035-8

Andreoni

(1989). Giving with impure altruism: Applications to charity and Ricardian equivalence. Journal of Political Economy , 97(6), 1447–1458.

Andreoni

, & Miller

J. H.

(1993). Rational cooperation in the finitely repeated prisoner’s dilemma: Experimental evidence. The Economic Journal , 103(418), 570–585. https://doi.org/10.2307/2234532

Andreoni

, & Varian

(1999). Preplay contracting in the prisoners’ dilemma. Proceedings of the National Academy of Sciences , 96(19), 10933–10938.

Batson

C. D.

, & Moran

(1999). Empathy-induced altruism in a prisoner’s dilemma. European Journal of Social Psychology , 29(7), 909–924.

Baumert

, Schlösser

, & Schmitt

(2014). Economic games: A performance-based assessment of fairness and altruism. European Journal of Psychological Assessment , 30(3), 178–192. https://doi.org/10.1027/1015-5759/a000183

Bereby-Meyer

, & Roth

A. E.

(2006). The speed of learning in noisy games: Partial reinforcement and the sustainability of cooperation. American Economic Review , 96(4), 1029–1042.

Bolton

G. E.

, & Ockenfels

(2000). ERC: A theory of equity, reciprocity, and competition. American Economic Review , 90(1), 166–193. https://doi.org/10.1257/aer.90.1.166

Camerer

C. F.

, & Hogarth

R. M.

(1999). The effects of financial incentives in experiments: A review and capital-labor-production framework. Journal of Risk and Uncertainty , 19(1/3), 7–42.

10.

Camerer

C. F.

, Loewenstein

, & Rabin

(Eds.). (2004). Advances in behavioral economics . Princeton University Press.

11.

Capraro

, Smyth

, Mylona

, & Niblo

G. A.

(2014). Benevolent characteristics promote cooperative behaviour among humans. PLoS One , 9(8), e102881. https://doi.org/10.1371/journal.pone.0102881

12.

Charness

, Gneezy

, & Rasocha

(2021). Experimental methods: Eliciting beliefs. Journal of Economic Behavior & Organization , 189, 234–256. https://doi.org/10.1016/j.jebo.2021.06.032

13.

Charness

, & Rabin

(2002). Understanding social preferences with simple tests. The Quarterly Journal of Economics , 117(3), 817–869. https://doi.org/10.1162/003355302760193904

14.

Chen

, & Zhang

(2021). Do elicited promises affect people’s trust?—Observations in the trust game experiment. Journal of Behavioral and Experimental Economics , 93, 101726. https://doi.org/10.1016/j.socec.2021.101726

15.

Chernyak

, Leimgruber

K. L.

, Dunham

Y. C.

, Hu

, & Blake

P. R.

(2019). Paying back people who harmed us but not people who helped us: Direct negative reciprocity precedes direct positive reciprocity in early development. Psychological Science , 30(9), 1273–1286.

16.

Cook

K. S.

, & Cooper

R. M.

(2003). Experimental studies of cooperation, trust, and social exchange. In Ostrom

, & Walker

J. W.

(Eds.), Trust and reciprocity: Interdisciplinary lessons for experimental research (pp. 209–244). Russell Sage Foundation.

17.

Cox

J. C.

, Sadiraj

, & Sadiraj

(2001). Trust, fear, reciprocity, and altruism . University of Arizona.

18.

Cummings

R. G.

, Harrison

G. W.

, & Rutstrom

E. E.

(1995). Homegrown values and hypothetical surveys: Is the dichotomous choice approach incentive-compatible? American Economic Review , 85(1), 260–266.

19.

Fehr

, & Fischbacher

(2002). Why social preferences matter–The impact of non-selfish motives on competition, cooperation and incentives. The Economic Journal , 112(478), C1–C33. https://doi.org/10.1111/1468-0297.00027

20.

Fehr

, & Schmidt

K. M.

(1999). A theory of fairness, competition, and cooperation. The Quarterly Journal of Economics , 114(3), 817–868. https://doi.org/10.1162/003355399556151

21.

Fischbacher

, Gächter

, & Fehr

(2001). Are people conditionally cooperative? Evidence from a public goods experiment. Economics Letters , 71(3), 397–404.

22.

Gintis

, Bowles

, Boyd

, & Fehr

(2003). Explaining altruistic behavior in humans. Evolution and Human Behavior , 24(3), 153–172.

23.

Greig

, & Bohnet

(2008). Is there reciprocity in a reciprocal-exchange economy? Evidence of gendered norms from a slum in Nairobi, Kenya. Economic Inquiry , 46(1), 77–83.

24.

Grimm

, & Mengel

(2009). Cooperation in viscous populations—Experimental evidence. Games and Economic Behavior , 66(1), 202–220.

25.

Hammerstein

(Ed.). (2003). Genetic and cultural evolution of cooperation . MIT Press.

26.

Hayashi

, Ostrom

, Walker

, & Yamagishi

(1999). Reciprocity, trust, and the sense of control: A cross-societal study. Rationality and Society , 11(1), 27–46. https://doi.org/10.1177/104346399011001002

27.

Hruschka

D. J.

, & Henrich

(2006). Friendship, cliquishness, and the emergence of cooperation. Journal of Theoretical Biology , 239(1), 1–15. https://doi.org/10.1016/j.jtbi.2005.07.006

28.

Janssen

M. A.

(2008). Evolution of cooperation in a one-shot Prisoner’s Dilemma based on recognition of trustworthy and untrustworthy agents. Journal of Economic Behavior & Organization , 65(3–4), 458–471.

29.

Jordan

J. J.

, Capraro

, & Rand

D. G.

(2014). Heuristics guide the implementation of social preferences in one-shot prisoner’s dilemma experiments. Scientific Reports , 4, Article 6790. https://doi.org/10.1038/srep06790

30.

Kahneman

, & Tversky

(1979). Prospect theory: An analysis of decision under risk. Econometrica , 47(2), 263–291.

31.

Kiyonari

, Tanida

, & Yamagishi

(2000). Social exchange and reciprocity: Confusion or a heuristic? Evolution and Human Behavior , 21(6), 411–427. https://doi.org/10.1016/S1090-5138(00)00055-6

32.

Levine

D. K.

(1998). Modeling altruism and spitefulness in experiments. Review of Economic Dynamics , 1(3), 593–622. https://doi.org/10.1006/redy.1998.0023

33.

Mangone

(2020). Beyond the dichotomy between altruism and egoism: Society, relationship, and responsibility . IAP.

34.

Martin

J. M.

, Gonzalez

, Juvina

, & Lebiere

(2014). A description–experience gap in social interactions: Information about interdependence and its effects on cooperation. Journal of Behavioral Decision Making , 27(4), 349–362. https://doi.org/10.1002/bdm.1810

35.

Morris

M. W.

, Sim

D. L.

, & Girotto

(1998). Distinguishing sources of cooperation in the one-round prisoner’s dilemma: Evidence for cooperative decisions based on the illusion of control. Journal of Experimental Social Psychology , 34(5), 494–512.

36.

Ostrom

, & Walker

(Eds.). (2003). Trust and reciprocity: Interdisciplinary lessons for experimental research . Russell Sage Foundation.

37.

Rabin

(1993). Incorporating fairness into game theory and economics. The American Economic Review , 83(5), 1281–1302.

38.

Rand

D. G.

, Dreber

, Ellingsen

, Fudenberg

, & Nowak

M. A.

(2009). Positive interactions promote public cooperation. Science , 325(5945), 1272–1275. https://doi.org/10.1126/science.1177418

39.

Read

(2005). Monetary incentives, what are they good for? Journal of Economic Methodology , 12(2), 265–276.

40.

Shafir

, & Tversky

(1992). Thinking through uncertainty: Nonconsequential reasoning and choice. Cognitive Psychology , 24(4), 449–474.

41.

Shaw

, Barakzai

, & Keysar

(2019). When and why people evaluate negative reciprocity as more fair than positive reciprocity. Cognitive Science , 43(8), e12773.

42.

Suneja

, & Das

(2022). Impact of social affinity on altruism: Experimental evidence from the Ultimatum Game. International Journal of Social Economics , 50(3), 335–344. https://doi.org/10.1108/IJSE-07-2022-0481

43.

Weiss

, & Peres

(2014). Beyond the altruism-egoism dichotomy: A new typology to capture morality as a complex phenomenon. In Jeffries

(Ed.), The Palgrave handbook of altruism, morality, and social solidarity . Palgrave Macmillan. https://doi.org/10.1057/9781137391865_4