Framing Zero, Why Losing Nothing is Better than Gaining Nothing

“Choices are easy when you have nothing to lose” -Barbara Delinsky

Introduction

Imagine being offered the following bet: 20% chance of getting $9 and 80% chance of getting nothing. Would it make any difference if zero was framed as losing nothing instead of gaining nothing? Furthermore, if you lost $5, would the same bet be more or less appealing?

The aim of this paper is to draw attention to the importance of how zero is framed in decision research. This fact needs to be recognized and managed in future studies, as zero framing has a significant impact on the results. Thus, the main contribution of this paper is to draw attention to the effect different frames of zero have on judgement and decision making. The second main contribution is to demonstrate the mechanisms by which the framing of zero effects judgement and choice. In a simple gamble context, gain nothing is perceived as negative and acts as a reference point that makes other parts of the bet feel worse, whereas lose nothing is perceived as positive and makes other parts of the bet feel better ( Wardley, M., & Alberhasky, M, 2021).

Conceptual Development and Theoretical Framework 

• Tversky and Kahneman (1986) and Bateman et al., (2007)

Tversky and Kahneman (1981) theorise that information is coded as positive or negative, depending on how choices are framed, and this coding affects value based on the value function of prospect theory. Bateman et al., (2007) find that a bet with a small chance of winning and a greater chance of winning nothing,  is judged as less attractive than a bet with a small chance of winning and a greater chance of losing. Since winning nothing is clearly better than losing a small amount of money, participants show a preference for the dominated options. The payoff combination (a certain amount of money, win nothing) is not evaluable, and adding a small amount of loss of money makes a certain amount of money ‘come alive with feelings’ and causes it to be weighted in the judgement of attractiveness (Bateman et al., 2007).

• Framing, Affect, and Reference Points

In general, many studies support the notion that the effect of message framing is highly dependent on emotional reactions. Kahneman & Frederick (2007) said that framing effects arise from an emotional response where sure gains are particularly attractive and sure losses are particularly unattractive.

According to prospect theory, losses and gains are defined as a neutral reference point or change from the status quo. However, reference points can come from a variety of sources, such as social comparisons and goals, and may not necessarily represent zero or the status quo (Higgins & Liberman, 2018). In particular, buyers can focus on their own personal utility of acquiring goods in determining their willingness to pay, and sellers can use market value as a reference point in determining their willingness to receive, thereby accounting for observed differences.

• Predictions

Wardley, M., & Alberhasky, M (2021) estimate that in the context of the two decision problems Tversky and Kahneman (1986) and Bateman et al., (2007) gain nothing has a negative effect whereas loss nothing has a positive effect. When zero is framed as nothing to lose, it induces a positive influence and draws attention to the potential profit of the bet, which is no worse than before playing the bet and potentially winning bet amount (Wardley, M., & Alberhasky, M., 2021).

Methodological 

• Study Replication and Extension of Tversky and Kahneman (1986)

There are two problems which are used by this study, in problem 3 the framing of zero is changed to lose nothing a greater proportion of participants will select the risk-seeking option whereas in problem 4 if the framing of zero is changed to gain nothing a greater proportion of participants will select the risk-averse option. In this study, the problem is modified, then made the comparison with the original one.  That is summarised as below. 

Tabel 1. Result of study Replication of Tversky and Kahneman

The result indicated, in condition 1 where the original study was conducted, the majority of participants selected the risk-averse option in problem 3 and the risk-seeking option in problem 4. However, by modifying the framing of zero in problem 3 significantly more participants selected the risk-seeking option, and in problem significantly more participants selected the risk-averse option in comparison to the original. This demonstrates that these decision problems are heavily dependent on how zero is framed. 

• Gain Frame Replication with Real Money

The purpose of this study is to replicate problem 3 with real instead of hypothetical payouts. This study utilised the study by Kühberger that framing effects tend to be weak for small payoffs. Further, the study employs two sets of conditions, where  in condition 1 participants faced a choice between A: “A sure gain of $5.00” or B: “A 50% chance to gain $10.00 and a 50% chance to gain nothing.” In condition 2 choice A was identical, but the framing of zero in option B was changed from “gain” to “lose. The results of a binomial test indicate significantly more participants selected the risk averse option in comparison to the risk-seeking option. The results of the current study replicate show that in the realm of gains, participants are risk-averse and more importantly how zero is framed affects choice with real payouts.

• Replication and Extension of Bateman et al., (2007)

In relation to the bet used by Bateman et al., (2007), lose $.05 will be perceived as more attractive than gain nothing but lose nothing will be perceived as more attractive than lose $.05. In conducting study, participants were asked to consider the prospect of a bet with a “20% chance to gain $9” and in condition 1 an “80% chance to gain nothing,” in condition 2 an “80% chance to lose nothing,” and in condition 3 an “80% chance to lose $.05.” 

The results of this study provide evidence that how zero is framed can affect the perceived attractiveness of a hypothetical gamble in an attribute framing problem,  which extends the founding of Bateman et al., (2007) by demonstrating how zero is framed affects the perceived attractiveness of a simple gamble. The result show, with a 20% chance to gain $9 and an 80% chance to lose $.05 is more attractive than the same bet paired with an 80% chance to gain nothing. Conversely, when there is an 80% chance to lose nothing this is more attractive than an 80% chance to lose $.05

• Affective Mechanism

Utilizing a similar methodology to Bateman et al., (2007),  this study tests the evaluability and reference point explanations by having participants provide affective ratings for each part of the bet. Specifically affect was assessed towards each amount ($9; gain nothing/lose nothing/lose $.05) and each probability (80%; 20%)

The evaluability and affect heuristic explanation predicts no significant differences on the affective ratings of the $9 between the gain nothing and lose nothing conditions, but a significant difference on ratings of the $9 between the lose nothing and lose $.05 conditions.

These predictions are not supported by the data. Thus, the differences in perceived bet attractiveness do not conform to an evaluability and affect heuristic explanation.This pattern of results explains why losing $.05 is judged as more attractive than gaining nothing, but losing nothing is judged as more attractive than losing $.05.

Conclusion 

Across four studies conducted, this paper  shows how zero is framed has a substantial effect on judgement and decision-making. In study replication from Tversky and Kahneman (1986), it shows the extent of two risky choice decision problems. The study replication with real money, the gain frame findings combined with real payouts. Further, replication of Bateman et al., (2007) extend an attribute framing problem from and explore the mediating role of affect. Study effective mechanisms built on Bateman by further exploring the affective mechanism and testing a reference point and affect heuristic explanation. Further, the paper shows that in the context of a simple gamble, losing a small amount is preferred over gaining nothing, but losing nothing is preferred over losing a small amount. 

Direview dari artikel jurnal : Marcus Wardley a, M. A. ( 9 November 2020). Framing zero: Why losing nothing is better than gaining nothing. Journal of Behavioral and Experimental Economics 90 (2021) 101641, 1-8.

Editor: Phylicia Febian, Nurul Sekararum, Daffa Dzakwan Jamal

Illustration by: Nadira Meuthia Jefri