Reducing Delay Discounting of
The Financially Deprived through Framing
Candrika Sagitasari MSc Marketing Management
Master Thesis MSc Marketing Management
Reducing Delay Discounting of
The Financially Deprived through Framing
University of Groningen Faculty of Economics and Business
MSc Marketing Management Master Thesis
First supervisor: Dr. Mehrad Moeini-Jazani Second supervisor: Prof. Dr. B. M. Fennis
Completion Date: January 13th, 2020
Candrika Sagitasari S3884538 Bloemstraat 41A
9712 LB, Groningen, The Netherlands +31627513925
ABSTRACT
In general, financially deprived people have a higher delay discounting than people under a better financial condition. This study aims to find the best solution to help financially deprived people to reduce their delay discounting in order to be able to make the best decision out of all available options. This study argues that by using framings, it can effectively decrease the delay discounting. Therefore, a 2 (financial status: financially deprived vs. financially non-deprived)
× 4 (framing: Traditional vs. SS Zero vs. LL Zero vs. Explicit Zero) study was conducted. In this study (n = 800), financial status and framings were manipulated and randomly assigned participants either to one out of eight conditions. Based on the results, even though the interaction between financial deprivation and asymmetrical zero framing was insignificant, the main effects were significantly influencing participant’s delay discounting. In overall, the findings suggest that SS Zero framing is the most effective framing in reducing delay discounting for financially deprived people by highlighting the opportunity cost of smaller, sooner reward, which makes them understand what would be the prospect to lose.
Keywords: delay discounting, financial deprivation, framing, explicit zero framing, SS zero
ACKNOWLEDGEMENT
This thesis is the peak of my academic journey, so far, starting from my bachelors in Jakarta, Indonesia and my masters, here in Groningen, Netherlands. One of the motivations for me to pursue a higher education abroad was to learn something new from the perspective of the other part of the world. I hope I can bring all of lessons I have learnt from both inside and outside classes back to Indonesia and be something useful, especially with poverty as the topic of my thesis, which is still a crucial area in Indonesia.
Firstly, I would like to thank Allah SWT for never gets tired listening to all of my endless prayers. I would like to thank my thesis supervisor, Dr. Mehrad Moeini-Jazani, for his guidance, helps, insights, and constructive critics. I would also like to thank my second supervisor, Prof. Dr. B. M. Fennis, for taking his time to evaluate my thesis.
TABLE OF CONTENTS
INTRODUCTION ...1
THEORETICAL FRAMEWORK ...2
The Condition of Poverty ...2
The Correlation between Poverty and Delay Discounting ...3
Opportunity Cost on Poverty ...5
Designing The Opportunity Cost Through Framing ...6
METHODOLOGY ... 12
Participant and Design ... 13
General Procedure ... 13
RESULTS ... 17
Data Inspection ... 17
Feeling of Financial Deprivation, Asymmetrical Zero Framing, and Delay Discounting ... 17
Control Check ... 20
Testing the Mediating Role of Thought Rating ... 21
Testing the Mediating Role of Affect ... 26
GENERAL DISCUSSION... 27
Limitation and Future Research ... 28
Implication ... 29
Conclusion ... 29
References ... 30
INTRODUCTION
Poverty is a problem that the world has been battling for a long time, yet it is still far from winning it. In 2015, there is 10 percent of the world’s population—736 million people—living in a condition where people are having less than $1.90 a day, a condition called as extreme poverty. For the last few years, the number is reflecting slow progress that makes World Bank uncertain on achieving 2030’s target to only have less than 3 percent of people living in extreme poverty (World Bank 2019). Why is it so hard to escape poverty? What makes a family lives in poverty through generations? Questions about poverty are abundant and have long kept philosophers, policy-makers, and social scientists busy (Shah, Mullainathan, and Shafir 2012), including the 2019’s winners of the Nobel Memorial Prize in Economic Science who received the honour because of their experimental approach to alleviating global poverty.
People who live in a financially deprived condition are usually behaving in ways that worsen the condition. Financially deprived leads to present-mindedness, a term where everything in the present is the most important while neglecting the future (Farah and Hook 2017; Shah, Mullainathan, and Shafir 2012). For instance, delay discounting, when they discount the future more than financially non-deprived people by choosing smaller, sooner rewards instead of waiting for larger, later rewards (Tanaka, Camerer, and Nguyen 2010).
THEORETICAL FRAMEWORK
The Condition of Poverty
Some of the most common problems from living in poverty are shorter life expectancy, higher risk of violence and crime, living in a bad neighbourhood, and minimal access to health care and formal credit markets (Banerjee and Duflo 2008; Haushofer and Fehr 2014). Additionally, children will have a higher tendency to get malnourished, have less opportunity to earn proper education by attending school, and have a probability to face similar life and financial forecasts to their parents (Azariadis and Stachurski 2004; Haushofer and Fehr 2014). Most importantly, financially deprived people are living with smaller cognitive resources compared to people under a wealthier condition because some problems loom larger that takes most of the brain capacity. The smaller cognitive resources will influence the cognitive function for making the best decision for all of the essential aspects of life (Shafir 2017).
To see from the financial perspective, living in poverty means not being able to afford fundamental needs to live the life while facing other endless financial difficulties (Shafir 2017). Among financially deprived people, it is common to dedicate 30 percent of their income only for housing expenses (Gennetian and Shafir 2015). If the rest of the income went for other fundamental needs, such as food and clothing, there would be nothing left to manage any insurances (Shafir 2017), money to be saved for the future (Barr 2012), children’s education fees, and pension fund (Mullainathan and Shafir 2013).
discouraged about life (Shafir 2017). Being seen as a society's burden, lazy, and unmotivated (Horan and Austin 1974; Rogers-Dillon 1995) can affect financially deprived people that eventually can lead them to underperform (Hall, Zhao, and Shafir 2014), including on missing the opportunity for significant benefits (Bissett and Coussins 1982). One of the fundamental steps to solve this condition is an appreciation of the power of context to reform their thought and behavior (Shafir 2017).
The Correlation between Poverty and Delay Discounting
People generally do delay discounting, which means they have a preference for immediate rewards over larger rewards that will occur in the future (Frederick, Loewenstein, and O’Donoghue 2002; Olivola and Wang 2016; Read, Frederick, and Scholten 2013). Additionally, people also do temporal discounting—tendency to perceive events occur in the distant future is less critical than events in the present or near future (Ainslie 2001). However, even though it is common for people to do delay discounting, financially deprived people tend to have a higher delay discounting compared to people in a wealthier condition (Haushofer and Fehr 2014).
(Bair 2005) and regularly play lotteries (Haisley, Mostafa, and Loewenstein 2008). Additionally, they consume high-calorie foods because of the lack of resources to practice and apply a healthy lifestyle, which can increase the tendency of getting diet-related diseases in the future (Cheon and Hong 2017).
The reasons behind the high delay discounting on the poverty context are varied. For instance, financially deprived people are leaning to be more risk-averse, lower willingness to spare, and invest in the current income in favor of higher future returns (Dohmen et al. 2011; Haushofer and Fehr 2014). They are weakening the use of self-control, which makes it tougher to make decisions favoring the future (Banerjee and Mullainathan 2010). Additionally, depression, anxiety disorder (WHO 2001), the dreadful feeling they generate from the continuous income shocks (Banerjee and Duflo 2008; Haushofer, Schunk, and Fehr 2013), feeling of powerlessness (Joshi and Fast 2013), and stress caused by the poverty condition make it harder to generate a high-quality decision on choosing the best option (Haushofer and Fehr 2014; Lerner, Li, and Weber 2013; Mullainathan and Shafir 2013).
Opportunity Cost on Poverty
An opportunity cost is the calculation of the highest values from the alternatives, options, or other opportunities that were rejected for choosing the other available one (Buchanan 2008). To make the most excellent decision, people should put opportunity cost into the account, but still, people are neglecting the opportunity cost (Shafir and Thaler 2006). Theoretically, people start to consider the opportunity cost only when they have limited financial resources (Spiller 2011) because that condition promotes trade-off thinking (Mullainathan and Shafir 2013).
Some believe the tendency to neglect the opportunity cost for financially deprived people is less likely to happen as they only have limited financial resources (Frederick et al. 2009). They identify opportunity cost or trade-off more salient than people under a wealthier condition (Shafir 2017). Additionally, it is specifically crucial for financially deprived people because their limited financial resources make them having only a small area to create any wrong decisions, and they also realize that money can only be spent once (Bertrand, Mullainathan, and Shafir 2006; Spiller 2011).
In reality, neglecting the opportunity cost still happens among financially deprived people because besides having limited financial resources, they also have limited cognitive resources to think about something that they consider as unimportant from their perspective (Harinck et al. 2007; Haushofer and Fehr 2014). With the limited cognitive capacity they own, they often fail to see and calculate thoroughly the alternative that is not presented (Frederick et al. 2009; Jones et al. 1998).
decision making process is by emphasizing the opportunity cost because it can change the perceived attractiveness of other available opportunities and lead them toward the best decision (Frederick et al. 2009; Harinck et al. 2007; Spiller 2011). For instance, as simple as reminding people that not making this particular purchase will make them have more money to be allocated for other purchases reduces their purchase probability (Frederick et al. 2009). However, to help financially deprived people, it is still unclear what is the best way on how to design the opportunity cost to be salient.
Designing The Opportunity Cost Through Framing
As most of the financially deprived people’s cognitive are occupied only for problems they consider as important, it leads them to ignore, discount, and forget about other aspects (Shafir 2017). For instance, based on (Mani et al. 2013), when financially deprived people were asked to think about their financial condition and then did some tests to measure their intelligence and cognitive control, the results of the tests were terrible compared to financially non-deprived people. To help them to get out of this behavior, their thinking process when making the decisions have to be changed to think more about the future and other unrealized essential aspects instead of just being extremely focused on present problems and neglecting the other (Shafir 2017).
monetary choices. Additionally, in the literature of delay discounting, some frames were successfully proven as effective in reducing delay discounting, as shown in Table 1.
Table 1. Framing and The Position of Opportunity Cost
Smaller, Sooner Opportunity Cost
Implicit Explicit Larger, Later Opportunity Cost Implicit Traditional (1): $100 today OR $150 in one year SS Zero (2):
$100 today and $0 in one year OR $150 in one year Explicit LL Zero (3): $100 today OR
$0 today and $150 in one year
Explicit Zero (4):
$100 today and $0 in one year OR
$0 today and $150 in one year
The first frame is the Traditional framing with hidden-zero by (Estle et al. 2007). This framing is allowing people to choose between the smaller-sooner (SS) reward and larger-later (LL) reward. Under this Traditional framing, people will perceive SS reward as a good sooner or immediate option and LL reward as a delayed with a better amount of reward option. The hidden-zero technique means framing with the absence of zero in both options. Zero has a role in explaining what would happen in the present time or future time, depending on its position, if people decided to choose a particular option. With no zeros in both options, this framing is not explicitly elaborating to the people that choosing a specific option also means choosing not to receive the other available option (see Table 1, part 1).
both options. Having zero in both options (i.e., next to SS and LL reward) makes people understand what would happen if they decided to choose a particular option. In other words, this framing makes opportunity costs from both options salient (see Table 1, part 4). In overall, research shows that people presented with Explicit Zero framing had less delay discounting compared to people presented with Traditional framing, even though both framings have logically identical options.
The reason why Explicit Zero framing is more effective than the Traditional framing is that people would prefer on choosing an option that has an improving sequence over time, an attribute that Traditional framing does not have because of the absence of the zero (Ariely and Zauberman 2003; Frederick, Loewenstein, and O’Donoghue 2002). More specifically, the zero in LL reward makes the option looks like having an improving sequence because as people read the LL reward, they understand that even though they would get nothing today, they would get a larger reward in the future. For instance, “I would get nothing today but $150 in one year” seems better than “I would get $100 today but nothing in one year”. This may suggest that making the opportunity costs only on the LL reward side of the trade-off (see Table 1, part 3) salient may be enough to reduce people’s delay discounting.
recent research has explored the effectiveness of SS Zero and LL Zero framing vis-à-vis with Traditional and Explicit Zero framing (Read, Olivola, and Hardisty 2017). Results highlight the existence of Asymmetric Subjective Opportunity Cost (ASOC) effect in people’s decisions, such that salience of zero on the SS reward (i.e., SS Zero framing) seems sufficient to reduce people’s delay discounting, similar to the Explicit Zero framing where zeros are salient on both sides of the trade-off (see Table 1, part 4). This happens because people already understand the opportunity costs of LL reward (not having SS reward) but tend to neglect the opportunity costs of SS reward (not having the LL reward). In other words, people are only considering the cost that occurs in the present time but not in the future. Thus, reminding the opportunity cost of SS reward will help them to not neglect it.
Figure 1. Mean Level of Patience as a Function of Framing, in Study 1 (Read, Olivola, &
To test which frame is more effective in reducing delay discounting, Read and colleagues (2017) compared all Traditional, SS Zero, LL Zero, and Explicit Zero framings. As the results in Figure 1 shows, by highlighting zero on the SS reward (SS Zero framing), it will make the opportunity cost of SS reward more salient than before. This technique will help people to reduce delay discounting by choosing LL rewards more, just like what they do in Explicit Zero framing. On the other hand, since the opportunity cost of LL reward has already been salient in their mind, putting zero on the LL reward (LL Zero framing) will not impact the delay discounting, just like in Traditional framing. In conclusion, the position of zero has a vital role as a reminder, and also it will make the opportunity cost seems more salient to people.
While these findings are enlightening, these framings have never been tested in the context of poverty. When tested into the general population, it shows that people are doing delay discounting by preferring SS rewards instead of LL rewards. Through one study, this paper predicts that framing will influence the number of LL reward chosen under a different feeling of financial deprivation (feeling of financially deprived and feeling of financially non-deprived). This study predicts that people under the feeling of financially deprived will be more delay discounting than people under financially non-deprived, but SS Zero framing will help them more than other framings because when it is presented with SS Zero framing, it will be clear for them see what would be the loss in the future if they did not choose the LL reward. Furthermore, for financially non-deprived people, SS Zero and Explicit Zero will perform better than LL Zero and Traditional framings. Therefore, the following hypotheses are developed:
H1: Framing will moderate the effect of poverty feelings on delay discounting such
is larger among the financially deprived that it is among the financially non-deprived.
H2: Among the financially deprived, SS Zero and Explicit Zero framings reduce
delay discounting compared to LL Zero and Traditional framings.
H3: Among the financially non-deprived, the effect of framing is less pronounced
and thus it is expected a small to null effect in the differences between SS Zero and Explicit Zero framings compared to LL Zero and Traditional framings.
Figure 2. Conceptual Model
( - )
Financial
Deprivation Discounting Delay Zero Framing
METHODOLOGY
The purpose of this study is to test the effectiveness of the four framings (i.e., Traditional, SS Zero, LL Zero, and Explicit Zero) in presenting intertemporal choices in reducing delay discounting of the financially deprived people. In doing so, we manipulate participant’s feeling of financial deprivation (i.e., financially deprived and financially non-deprived) and four different framings. In overall, the aim of this study is to find the best solution to help financially deprived people to get out of their financial problems by being more future-oriented.
G*Power was used to determine the sample size for this study (Faul et al. 2009) to have a power of 0.80 and an α-error probability of 0.05 to detect the hypothesized effect. For this study, power analysis for an ANOVA generated a minimum sample of 762 to detect a small-sized two-way interaction effect between the manipulation of the feeling of financial deprivation and framing on delay discounting. When the data was collected, the sample size exceeded the minimum sample size in a single attempt.
Participant and Design
Eight hundred and seventy American participants took part in a 2 (financial status: deprived vs. non-deprived) × 4 (framing: Traditional, SS Zero, LL Zero, and vs. Explicit Zero) study.
General Procedure
At the beginning of the survey, participants needed to do Captcha and a control question as a first requirement to proceed with the survey. Next, participants were informed about the survey in a brief (i.e., it contains some brief writing tasks, it will take around 10 - 15 minutes, and do not leave the device until the survey finished) without explaining the actual means of the survey.
In the next part, participants needed to answer some questions about their age, gender, number of people living in their household, including themselves, and which state in America they were living in at that moment. After answering those questions, it continued with the manipulation feeling of financial deprivation where participants were randomly assigned either to financially deprived or financially non-deprived conditions (Moeini-Jazani, Albalooshi, and Seljeseth 2019). Briefly, the procedure of this manipulation begun with a question about participant’s monthly discretionary income using a sliding bar continued with false feedback on participant’s financial status, and lastly, a brief writing task.
had a sliding bar ranging from $0 to $2,000 (or more). The purpose of providing different sliding bar was to make participants in the financially deprived condition with a maximum of $50,000 sliding bar should experience a relative feeling of financial deprivation because the range was extensive. On the contrary, participants in the financially non-deprived condition with a maximum of $2,000 sliding bar should experience a relative feeling of financial stability and adequacy.
After indicating their monthly discretionary income, participants in both conditions received information that a system will compare and calculate their relative financial status with other participants who have a similar profile with them from the registered data. To make it looked like a real system that was calculating the data, after reading the information, participants saw a loading sign for five seconds. In reality, there was no system at all. Next, participants received false feedback based on their assigned condition (see Appendix A) and asked to write about how they feel after knowing their financial status.
After the feeling of financial deprivation manipulation, participants entered a manipulation check part to check whether the manipulation beforehand was successful or not. In this part, participants need to answer four questions (e.g., “To what extent do you feel financially constrained?” and “How would you rate your financial position in comparison to your peers?”) with a 7-point Likert scale (see Appendix B).
two choices. The choices were presented as below, expressed in four different framings, that was used in this study:
Traditional: [A] $14 today, OR [B] $25 in 19 days
SS Zero: [A] $14 today and $0 in 19 days, OR [B] $25 in 19 days LL Zero: [A] $14 today, OR [B] $0 today and $25 in 19 days
Explicit Zero: [A] $14 today and $0 in 19 days, OR [B] $0 today and $25 in 19 days
In total, they were asked 18 different monetary choices (Read, Olivola, and Hardisty 2017), with eleven subsets questions from (Magen, Dweck, and Gross 2008), and seven subsets questions from (Kirby, Petry, and Bickel 1999), and one additional control question. The order of the 18 and one additional control questions are as shown in Table 2.
Table 2. The Amounts and The Delays for 11 Subsets from “Magen Items”, 7 Subsets from
“Kirby Items”, and 1 Control Item
Order Reward Delay Status Today Later Days
18 $4.10 $7.50 20 Magen items 19 $34 $50 30 Kirby items
After the participants had answered all of the monetary choices, they continued to the thought rating part. In this part, using 7-point Likert scale, participants indicated to what extent the six given thoughts (e.g., “I thought receiving any money now would be more valuable to me than in the future”) went through their minds when they made the monetary choices. Furthermore, the six thoughts were categorized into four groups; desire for immediate reward, consideration of future gain, consideration of future loss, and consideration waiting time (see Appendix C).
Participants then continued to the next part, affect. They were asked to indicate their mood while they were doing the survey. They needed to answer six different bipolar mood questions (i.e., negative to positive, sad to happy, stressed to relaxed, anxious to calm, aroused to still, and bad to good).
RESULTS
Data Inspection
From 870 participants collected, 70 participants excluded before analyzing the data. Excluded participants were either failed on the first two questions of the survey, which were Captcha and attention check questions (n = 19), did copy and paste on the writing task (n = 37) or were responding questions with unrelated answers (n = 14). Therefore, for the analysis, only 800 participants were included (Mage = 33.59, SD = 5.99; 464 females), which is more than the minimum sample size generated from G*Power for this study to test the hypothesized interaction between financial deprivation and asymmetrical zero framing on delay discounting.
Feeling of Financial Deprivation, Asymmetrical Zero Framing, and Delay Discounting
For this study, a 2 (financial status: deprived vs. non-deprived) × 4 (framing: Traditional vs. SS Zero vs. LL Zero, and vs. Explicit Zero) between-subjects ANOVA on participant’s delay discounting (number of LL rewards chosen) was conducted. Based on the analysis, it showed a significant main effect of financial status (F(792) = 16.087, p < .001, η2p = .020) and framing (F(792) = 19.436, p < .001, η2p = .069), but not significant effect on the interaction between financial status and framing (F(792) = .324, p = .808, η2p = .001). These findings showed that only the main effects were significant and did not find any interaction, as what would have been expected in H1.
participants who assigned to Explicit Zero (M = 8.27, SD = 5.65) and LL Zero (M = 5.97, SD = 5.76), they were different (t(792) = 2.99, p = .003). Lastly, those who assigned to LL Zero framing (M = 5.97, SD = 5.76) and Traditional framing (M = 5.55, SD = 5.55; t(792) = 0.41, p = .68) had no difference. These findings, therefore, supported H2 as below:
SS Zero ≈ Explicit Zero > LL Zero ≈ Traditional
Furthermore, in the financially non-deprived condition the pattern is similar, participants who assigned to SS Zero framing (M = 10.85, SD = 6.25) showed no difference to those who assigned to Explicit Zero (M = 9.71, SD = 5.86; t(792) = 1.33, p = .18). In comparison between Explicit Zero (M = 9.71, SD = 5.86) and LL Zero framing (M = 8.10, SD = 5.90), they were marginally different (t(792) = 1.92, p = .056). Lastly, those who assigned to LL Zero framing (M = 8.10, SD = 5.90) and Traditional framing (M = 7.29, SD = 5.21; t(792) = .96, p = .34) showed no difference. In overall, these findings replicate past findings that SS Zero and Explicit Zero framings are similarly effective in reducing delay discounting compared to LL Zero and Traditional framings, which both are equally ineffective in doing so (Read, Olivola, and Hardisty 2017).
financially deprived. Next, in the LL Zero framing, participants who felt financially non-deprived (M = 8.10, SD = 5.90) showed less delay discounting than those who felt financially deprived (M = 8.27, SD = 5.65; t(792) = 2.73, p = .006), similar to the pattern observed under the Traditional framing. Lastly, in the Explicit Zero framing, participants who felt financially non-deprived (M = 9.71, SD = 5.86) also showed marginally less delay discounting than those who felt financially deprived (M = 8.27, SD = 5.65; t(792) = 1.74, p = .08), similar to the pattern observed under the SS Zero framing. In overall, these findings show that SS Zero and Explicit Zero framing are equally effective in reducing delay discounting of the participants in financially deprived condition. Moreover, across all framings, the financially non-deprived participant’s delay discounting was less influenced by framing which also supported H3 (see
Figure 3).
Figure 3. Number of LL Reward Chosen in All Conditions
Finally, these results showed how Explicit Zero framing improves the future-oriented thinking among the financially deprived. Considering the similarities that were found between Explicit
8,27 5,87 9,71 5,55 9,71 8,10 10,85 7,29
Explicit Zero LL Zero SS Zero Traditional
Number of LL Reward Chosen
Zero and SS Zero framings, it follows that when the prospect of future loss in SS Zero became salient for the financially deprived (i.e., SS Zero framing: [A] $14 today and $0 in 19 days, OR [B] $25 in 19 days), they became future-oriented. However, making the prospect of future gains salient (i.e., LL Zero framing: [A] $14 today, OR [B] $0 today and $25 in 19 days), by presenting a merely increasing sequence in the LL reward, did not improve future-oriented thinking of the financially deprived.
Control Check
Another ANOVA conducted to test whether the significant levels of main effects and the interaction between financial status and framing (Model 1) remain the same after controlling for the participant's demographic characteristics as covariates or not. In Model 1, the size of the samples was 800 participants. However, for Model 2 and Model 3, two participants were excluded from the analysis because these participants were not classified their gender as male or female. Model 2 treated age, gender, and income as covariates, while Model 3 has other additional covariates, which were household size, level of education, employment status, and ethnicity.
Table 3. Control Check
Model 1 Model 2 Model 3
Variable b SEb t b SEb t b SEb t
Intercept 7.286 .602 12.098*** 7.102 .633 11.224*** 5.834 .756 7.729*** Financial Statusa -1.735 .801 -2.165** -1.667 .797 -2.092** -1.576 .791 -1.993** Explicit Zerob 2.429 .852 2,852** 2.681 .849 3.159** 2.640 .843 3.131** LL Zerob .81 .854 ,959 .834 .839 .994 .884 .832 1.062 SS Zerob 3.560 .852 4,181*** 3.617 .849 4.262*** 3.695 .843 4.383*** Financial Statusa x Explicit Zerob .291 1.155 .251 .330 1.146 .288 .217 1.139 .191 Financial Statusa x LL Zerob -.494 1.144 -.432 -.277 1.134 -.244 -.402 1.127 -.357 Financial Statusa x SS Zerob .598 1.142 .523 .735 1.136 .647 .618 1.128 .548 Age .024 .019 1.273 .012 .019 .624 Gender .053 .413 .128 .139 .413 .337 Income .227 .059 3.844*** .163 .065 2.517** Household Size -.070 .118 -.592 Level of Education .365 .176 2.071** Employment Status .234 .451 .51 Ethnicity 1.571 .431 3.647***
a = Non-Deprived condition as the base, b = Indicates dummy compared with the Traditional framing as the base.
*p < 0.10, **p < 0.05, ***p < 0.01
Testing the Mediating Role of Thought Rating
The first thought rating tested was the desire for immediate reward. In order to test it, a 2 (financial status: deprived vs. non-deprived) × 4 (framing: Traditional vs. SS Zero vs. LL Zero, and vs. Explicit Zero) between-subjects ANOVA on participant’s desire for immediate reward was conducted. The result showed a significant effects on main effect of financial status (F(792) = 14.165, p < .001, η2p = .018), and framing (F(792) = 5.023, p = .002, η2p = .019), but a non-significant interaction between financial status and framing (F(792) = .309, p = .819, η2p = .001). These results indicated that both financial status and framing were influencing participant’s thoughts on the desire for immediate reward (see Figure 4).
Figure 4. Number of Desire for Immediate Reward
The second thought rating tested was consideration of future gain. To test it, a 2 (financial status: deprived vs. non-deprived) × 4 (framing: Traditional vs. SS Zero vs. LL Zero, and vs. Explicit Zero) between-subjects ANOVA on participant’s consideration of future gain was conducted. The result showed a significant effect of the main effect of financial status (F(792) = 8.955, p = .003, η2p = .011), framing (F(792) = 2.980, p = .031, η2p = .011), but a
non-4.080
4.571
3.982
4.822
3.670 3.989 3.549 4.066
Explicit Zero LL Zero SS Zero Traditional
significant effect on the interaction between financial status and framing (F(792) = .826, p = .480, η2p = .003). These results indicated that both financial status and framing were influencing participant’s thoughts on consideration of future gain (see Figure 5).
Figure 5. Number of Consideration of Future Gain
The third thought rating tested was consideration of future loss. To test it, a 2 (financial status: deprived vs. non-deprived) × 4 (framing: Traditional vs. SS Zero vs. LL Zero, and vs. Explicit Zero) between-subjects ANOVA on participant’s consideration of future loss was conducted. The result showed no main effect of financial status (F(792) = .877, p = .349, η2p = .001), but a significant main effect of framing (F(792) = 3.176, p = .025, η2p = .012). However, the two-way interaction between financial status and framing was not significant (F(792) = .733, p = .533, η2p = .003). These results indicated that only framing influenced participant’s thoughts on consideration of future loss (see Figure 6).
5.630 5.400 5.755 5.178 5.813 5.989 5.912 5.593
Explicit Zero LL Zero SS Zero Traditional
Figure 6. Number of Consideration of Future Loss
The last thought rating tested was consideration of waiting time. In order to test it, a 2 (financial status: deprived vs. non-deprived) × 4 (framing: Traditional vs. SS Zero vs. LL Zero, and vs. Explicit Zero) between-subjects ANOVA on participant’s consideration of waiting time was conducted. Based on the analysis, it showed that the main effect of financial status was not significant (F(792) = .207, p = .650, η2p = .000), but the main effect of framing was significant (F(792) = 3.139, p = .025, η2p = .012). However, the interaction between financial status and framing was not significant (F(792) = 1.045, p = .372, η2p = .004). These results indicated that only framing influenced the participant’s thoughts on consideration of waiting time (see Figure
7).
Based on these results, none of the interaction effects were significant. It concluded that thought rating was not a mediator for financial status and framing. Additionally, a 2 (financial status: deprived vs. non-deprived) × 4 (framing: Traditional vs. SS Zero vs. LL Zero, and vs. Explicit Zero) between-subjects ANOVA on participant’s delay discounting with thought rating as a covariate was conducted. Based on the results, it showed that from four groups of
21.250 16.495 19.755 22.593 18.890 18.234 23.615 25.165
Explicit Zero LL Zero SS Zero Traditional
thought rating, only consideration of future loss that was not significant (F(788) = .503, p = .478, η2p = .001), while desire for immediate rewards (F(788) = 344.446, p > .001, η2p = .304), consideration of future gain (F(788) = 151.171, p > .001, η2p = .161), and consideration of waiting time (F(788) = 5.347, p = .021, η2p = .007) were all significant.
Figure 7. Number of Consideration of Waiting Time
The result of this analysis also showed that the main effect of financial status was not significant (F(788) = 2.666, p = .103, η2p = .003), main effect of framing was significant (F(788) = 17.072, p > .001, η2p = .061), and the interaction between financial status and framing was not significant (F(788) = .133, p = .940, η2p = .001). Based on these results, only desire for immediate reward, consideration of future gain, and consideration of waiting time that can influence participants’ delay discounting but this effect was largely independent from the financial status and framing.
18.560 25.486 24.018 20.000 18.648 22.213 25.835 24.176
Explicit Zero LL Zero SS Zero Traditional
Testing the Mediating Role of Affect
Another possible mediator tested was affect. In order to test it, a 2 (financial status: deprived vs. non-deprived) × 4 (framing: Traditional vs. SS Zero vs. LL Zero vs. Explicit Zero) between-subjects ANOVA on participant’s affect was conducted. The result showed only a significant main effect on financial status was insignificant (F(792) = 2.961, p = .086, η2p = .004). The main effect on framing (F(792) = 1.961, p = .118, η2p = .007) and the interaction between financial status and framing were not significant (F(792) = .504, p = .680, η2p = .002).
Moreover, a 2 (financial status: deprived vs. non-deprived) × 4 (framing: Traditional vs. SS Zero vs. LL Zero vs. Explicit Zero) between-subjects ANOVA on participant’s delay discounting with affect as covariate was conducted. The result indicated that the effect of affect was not significant (F(791) = .163, p = .686, η2p = .000), but the main effect of financial status was significant (F(791) = 15.814, p > .001, η2p = .020), and framing was significant (F(791) = 19.458, p > .001, η2p = .069). Lastly, the interaction between financial status and framing remained non-significant (F(791) = .332, p = .802, η2p = .001). Based on these results, it showed that affect did not influence on participant’s delay discounting.
GENERAL DISCUSSION
People, in general, are delay discounting, but financially deprived people have a higher tendency to do delay discounting than people under wealthier conditions. One of the solutions on how to reduce delay discounting is through framing. What framing does is based on the fact that people are putting the weight of the opportunity cost differently. They are very aware of the cost they might suffer now for choosing larger, later reward but put less attention to the opportunity cost of choosing a smaller, sooner reward, which results in receive nothing in the future. To help the financially deprived people to think more about the future, they need to see a reminder of the opportunity cost while making a decision.
Based on the results, it becomes clear that the most effective framing to reduce delay discounting of the financially deprived was a frame that was only highlighting the opportunity cost of choosing smaller, sooner reward, which was SS Zero framing. This framing highlights the consideration of future loss. It has been found that when financially deprived are reminded of this otherwise neglected opportunity cost, it helped them to see what would be the prospect to lose, which reduce their delay discounting. On the other hand, when the opportunity cost of choosing larger, later reward was highlighted (LL Zero), it was not affecting much because without reminding them, suggesting that opportunity cost had already salient for them or that the prospect of future gain is not as motivating as the salience of future loss.
what was not expected was that framing also helped financially non-deprived people, although to a smaller extent.
Limitation and Future Research
There are some limitations to this paper that found after analyzing the collected data. First, the number of participants was too small to be able to find the interaction between the feeling of financial deprivation and framing. Second, there was a lot of collected data (70 data) that had to excluded from the analysis due to the consequences of using MTurk. Some of these excluded data were bots with unrelated answers and abnormal pattern of responses(i.e., choose the first option available for all questions). Moreover, since participants had to use a device (i.e., computer or smartphone) to conduct the survey, for those who only wanted to get paid without giving serious effort on answering the questions, once they saw the brief writing task, they copy-pasted the answer from other sources.
Third, another probability of why the interaction was not significant can be caused by the uneven demographic of MTurk users in general. Even though the majority of MTurk users are well-educated, which makes them excellent in reading the instruction, they also have a low-income background and more likely to be depressed or anxious (Arditte et al. 2016; Berinsky, Huber, and Lenz 2012).
Implication
For the academic implication, these results are adding a new contribution, especially to the literature about delay discounting and poverty. Past papers have successfully designed framings to reduce delay discounting but never in the context of poverty (Magen, Dweck, and Gross 2008; Read, Olivola, and Hardisty 2017). Additionally, for managerial implication, these results can be insightful for some industries and companies—for instance, bank, debt-collector, insurance, or education industries. In order to make financially deprived people willing to save for their future, pay their debts, join any insurance, or send their children to school or university, they have to highlight the future loss and make it salient. In other word, these institutions should explain to them what would happen in the future, in detail, if they did not choose the most beneficial option.
Conclusion
References
Ainslie, G (2001), Breakdown of Will, New York: Cambridge University Press.
Ariely, Dan and Gal Zauberman (2003), “Differential partitioning of extended experiences,” Organizational Behavior and Human Decision Processes, 91 (2), 128–39.
Azariadis, C and J Stachurski (2004), “Poverty Traps,” Handbook of Economic Growth, 1 (Part A), 295–384.
Bair, S (2005), “Low-Cost Payday Loans: Opportunities and Obstacles,” Baltimore.
Banerjee, A and S Mullainathan (2010), “The Shape of Temptation: Implications for the Economic Lives of the Poor,” National Bureau of Economic Research.
Banerjee, Abhijit V. and Esther Duflo (2008), “What is middle class about the middle classes around the world?,” Journal of Economic Perspectives, 22 (2), 3–28.
Barr, Michael S (2012), No Slack : The Financial Lives of Low-Income Americans, Brookings Institution Press.
Baum, Andrew, J. P. Garofalo, and Ann Marie Yali (1999), “Socioeconomic status and chronic stress. Does stress account for SES effects on health?,” Annals of the New York Academy of Sciences, 896, 131–44.
Behrend, Tara S., David J. Sharek, Adam W. Meade, and Eric N. Wiebe (2011), “The viability of crowdsourcing for survey research,” Behavior Research Methods, 43 (3), 800–813. Bertrand, Marianne, Sendhil Mullainathan, and Eldar Shafir (2006), “Behavioral economics
and marketing in aid of decision making among the poor,” Journal of Public Policy and Marketing, 25 (1), 8–23.
Bissett, L and J Coussins (1982), Badge of Poverty: A New Look at the Stigma Attached to Free School Meals, Child Poverty Action Group.
Buhrmester, Michael D., Sanaz Talaifar, and Samuel D. Gosling (2018), “An Evaluation of Amazon’s Mechanical Turk, Its Rapid Rise, and Its Effective Use,” Perspectives on Psychological Science, 13 (2), 149–54.
Cannon, Christopher, Kelly Goldsmith, and Caroline Roux (2019), “A Self-Regulatory Model of Resource Scarcity,” Journal of Consumer Psychology, 29 (1), 104–27.
Chandler, Jesse and Danielle Shapiro (2016), “Conducting Clinical Research Using Crowdsourced Convenience Samples,” Annual Review of Clinical Psychology, 12 (1), 53– 81.
Cheon, Bobby K. and Ying Yi Hong (2017), “Mere experience of low subjective socioeconomic status stimulates appetite and food intake,” Proceedings of the National Academy of Sciences of the United States of America, 114 (1), 72–77.
Dohmen, T, D Huffman, S Jürgen, A Falk, U Sunde, and G G Wagner (2011), “Individual Risk Attitudes: Measurement, Determinants, and Behavioral Consequences,” Journal of the European Economic Association, 9 (3), 522–50.
Estle, Sara J., Leonard Green, Joel Myerson, and Daniel D. Holt (2007), “Discounting of monetary and directly consumable rewards: Research article,” Psychological Science, 18 (1), 58–63.
Farah, Martha J. and Cayce J. Hook (2017), “Trust and the poverty trap,” Proceedings of the National Academy of Sciences of the United States of America, 114 (21), 5327–29. Faul, Franz, Edgar Erdfelder, Axel Buchner, and Albert-George Lang (2009), “Statistical
power analyses using G*Power 3.1: Tests for correlation and regression analyses,” Behavior Research Methods, 41 (4), 1149–60.
Fiske, Susan T (2010), “Interpersonal Competition Can Cause Individuating Processes,” American Psychologist, 65 (8), 698–706.
time preference: A critical review,” Journal of Economic Literature, XL (June), 351–401. ———, Nathan Novemsky, Jing Wang, Ravi Dhar, and Stephen Nowlis (2009), “Opportunity
Cost Neglect,” Journal of Consumer Research, 36 (4), 553–61.
Gennetian, Lisa A and Eldar Shafir (2015), “The persistence of poverty in the context of financial instabililty: a behavioral perspective,” Journal of Policy Analysis and Management, 34 (4), 904–936.
Haisley, Emily, Romel Mostafa, and George Loewenstein (2008), “Myopic risk-seeking: The impact of narrow decision bracketing on lottery play,” Journal of Risk and Uncertainty, 37 (1), 57–75.
Hall, Crystal C., Jiaying Zhao, and Eldar Shafir (2014), “Self-Affirmation Among the Poor: Cognitive and Behavioral Implications,” Psychological Science, 25 (2), 619–25.
Harinck, Fieke, Eric Van Dijk, Ilja Van Beest, and Paul Mersmann (2007), “When gains loom larger than losses: Reversed loss aversion for small amounts of money,” Psychological Science, 18 (12), 1099–1105.
Hauser, David J. and Norbert Schwarz (2016), “Attentive Turkers: MTurk participants perform better on online attention checks than do subject pool participants,” Behavior Research Methods, 48 (1), 400–407.
Haushofer, Johannes and Ernst Fehr (2014), “On the psychology of poverty,” Science, 344 (6186), 862–67.
———, D Schunk, and Ernst Fehr (2013), “Negative Income Shocks Increase Discount Rates,” University of Zurich Working Paper.
Horan, Patrick M and Patricia L Austin (1974), “The Social Bases of Welfare Stigma,” Social Problems, 21 (5), 648–57.
Proceedings of the National Academy of Sciences of the United States of America, 114 (21), 5401–6.
Jones, Steven K., Deborah Frisch, Tricia J. Yurak, and Eric Kim (1998), “Choices and opportunities: Another effect of framing on decisions,” Journal of Behavioral Decision Making, 11 (3), 211–26.
Joshi, Priyanka D. and Nathanael J. Fast (2013), “Power and Reduced Temporal Discounting,” Psychological Science, 24 (4), 432–38.
Kahneman, Daniel and Amos Tversky (2000), Choices, Values, and Frames, Cambridge University Press.
Keith, Melissa G., Louis Tay, and Peter D. Harms (2017), “Systems perspective of amazon mechanical turk for organizational research: Review and recommendations,” Frontiers in Psychology, 8 (AUG).
Kirby, Kris N., Nancy M. Petry, and Warren K. Bickel (1999), “Heroin addicts have higher discount rates for delayed rewards than non-drug-using controls,” Journal of Experimental Psychology: General, 128 (1), 78–87.
Lerner, Jennifer S., Ye Li, and Elke U. Weber (2013), “The Financial Costs of Sadness,” Psychological Science, 24 (1), 72–79.
Magen, Eran, Carol S Dweck, and James J Gross (2008), “The hidden zero effect: Representing a single choice as an extended sequence reduces impulsive choice,” Asscociation for Psychological Science, 19 (7), 648–49.
Mani, Anandi, Sendhil Mullainathan, Eldar Shafir, and Jiaying Zhao (2013), “Poverty impedes cognitive function,” Science, 341 (6149), 976–80.
Mullainathan, Sendhil and Eldar Shafir (2013), Scarcity: Why Having Too Little Means So Much, New York: Times Books.
Olivola, Christopher Y. and Stephanie W. Wang (2016), “Patience auctions: the impact of time vs. money bidding on elicited discount rates,” Experimental Economics, 19 (4), 864–85. Read, Daniel, Shane Frederick, and Marc Scholten (2013), “Drift: An Analysis of Outcome
Framing in Intertemporal Choice,” Journal of Experimental Psychology: Learning Memory and Cognition, 39 (2), 573–88.
———, Christopher Y. Olivola, and David J. Hardisty (2017), “The value of nothing: Asymmetric attention to opportunity costs drives intertemporal decision making,” Management Science, 63 (12), 4277–97.
Rogers-Dillon, Robin (1995), “The Dynamics of Welfare Stigma,” Qualitative Sociology, 18 (4), 439–56.
Shafir, Eldar (2017), “Decisions in poverty contexts,” Current Opinion in Psychology, 18, 131–36.
——— and Richard H. Thaler (2006), “Invest now, drink later, spend never: On the mental accounting of delayed consumption,” Journal of Economic Psychology, 27 (5), 694–712. Shah, Anuj K, Sendhil Mullainathan, and Eldar Shafir (2012), “Some Consequences of Having Too Little,” American Association for the Advancement of Science, 338 (6107), 682–85. Shurtleff, D S (2009), “Improving savings incentives for the poor,” National Center for Policy
Analysis Brief Analysis, (672), 2009.
Spiller, Stephen A. (2011), “Opportunity Cost Consideration,” Journal of Consumer Research, 38 (4), 595–610.
Tomm, B and J Zhao (2018), “Scarcity Biases Attention to Motivationally Relevant Distractors,” Journal of Vision, 18 (10).
Tversky, Amos and Daniel Kahneman (1991), “Loss Aversion in Riskless Choice : A Reference-Dependent Model Author,” The Quarterly Journal of Economics, 106 (4), 1039–61.
WHO (2001), “World Health Report 2001 : Mental Health: New Understanding, New Hope,” Geneva.
Appendices
Appendix A — False Feedback
Feedback for financially deprived condition:
Our online calculator compared your information and discretionary income with a large, representative database of the US citizens who have a similar profile as you do. Based on the information you provided, our calculator identified you as an individual who currently is financially constrained, relative to similar others. That is, someone who may experience financial difficulties and, relatively lack
adequate financial resources (i.e., money). We would like to know more about
this from your perspective.
Please consider factors that limit your financial resources in daily life (e.g., mortgage or rent, daily or monthly expenses, lack of income, unexpected costs, etc.). Please take a few minutes to reflect and write about how it feels to be in a
relatively constrained financial position and to know that, on average, you may not have sufficient money to spend at your will or when unexpected
expenditures occur in daily life, compared to others who have sufficient
disposable income.”
Please take a few minutes to write your reflections and feelings about the
consequences of not having enough money and factors that limit your money to live a stable life in the space provided below. Please be as detailed and specific as possible.
Feedback for financially deprived condition:
Our online calculator compared your information and discretionary income with a large, representative database of the US citizens who have a similar profile as you do. Based on the information you provided, our calculator identified you as an individual who currently is financially adequate, relative to similar others; that is, someone who has a stable and sufficient amount of financial resources (i.e.
money) relative to others. We would like to know more about this from your
perspective.
Being financially stable means that you are able to spend less than you earn and able to pay for the basics of living while still having money set aside for any emergencies. We would like you to reflect and write about how it feels to
be financially stable and to know that you have sufficient money to use at your
will or when unexpected expenditures occur in daily life, relative to other
people who do not.
Appendix B — Manipulation Check
1. To what extent do you feel financially constrained?
2. How would you rate your financial position in comparison to your peers?
3. How would you rate your ability to spend money freely compared to your peers’ ability to spend money freely?
4. Overall, how would you rate your financial satisfaction at this moment?
Appendix C — Thought Rating
Desire for immediate reward:
1. When making choices, I thought receiving any money now would be more valuable to me than in the future.
2. When making choices, prospect of gaining more money in the future was attractive to me.
Consideration of future loss
1. When making choices, I thought the prospect of receiving less money in the future was unpleasant.
2. When making choices, the prospect of receiving nothing in the future concerned me.
Consideration of waiting time:
1. When making choices, I thought waiting to receive more money in the future was worthwhile.
