No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
Consistency as a Signal of Skills
Abstract
This paper studies the role of consistency as a signaling device. We propose a two-period model that highlights the informativeness of consistency as a signal of skills and allows for the analysis of consequences for behavior. In a simple principal–agent experiment, we test the basic intuition of the model. We show that consistency is indeed associated with skills. Consequently, consistency is valued by others, inducing people to act consistently.
Data, as supplemental material, are available at http://dx.doi.org/10.1287/mnsc.2016.2459.
This paper was accepted by Uri Gneezy, behavioral economics.
... For example, it has been argued that accumulating fatigue over a sequence of independent cases leads parole judges to be increasingly more harsh as they make more decisions concerning inmates (Danziger et al., 2011), food health inspectors to be increasingly less stringent as they visit more food establishments (Ibanez & Toffel, 2020), and journal editors to be increasingly more positive as they decide concerning more initial submissions (Orazbayev, 2017). 1 Although fatigue can drive serial position effects, we argue that they often emerge independently of fatigue. We suggest that the motivation of decision-makers to be consistent across and within sequences of decisions (Falk & Zimmermann, 2017;Haubensak, 1992) can be an important driver of many real-world serial position effects. Motivational accounts of serial position effects are rarely considered in organizational contexts, but they can have important practical consequences. ...
... Motivation for consistency may stem from a direct preference for consistency, for example, so as to avoid cognitive dissonance (Yariv, 2002), or from the fact that consistency violations are commonly perceived as signals of low skill and reliability (Falk & Zimmermann, 2017). We propose that the motivation to be or appear consistent can manifest in several ways, leading to serial position effects based on three distinct mechanisms: calibration (mentioned above) that leads to increases in extreme decisions, quotas that lead to increases in decisions restricted under quota, and strategic uncertainty that leads to increases in decisions that are likely to be scrutinized. ...
... In such settings, unlike those that we consider in this article, an implicit comparison among items in the sequence is almost inevitable, and evaluations are likely not (and arguably need not be) truly independent. However, decision-makers may strive to produce internally consistent sequences of decisions even when the items in the sequence are not later compared and ranked (Falk & Zimmermann, 2017;Haubensak, 1992). Thus, as long as decision-makers need to calibrate an evaluation scale (i.e., they have some initial uncertainty concerning the range of quality of items in the sequence), they may still strategically avoid early extreme judgments, making extreme decisions more likely in later serial positions. ...
Experts and employees in many domains make multiple similar but independent decisions in sequence. Often, the serial position of the case in the sequence influences the decision. Explanations for these serial position effects focus on the role of decision-makers' fatigue, but these effects emerge also when fatigue is unlikely. Here, we suggest that serial position effects can emerge due to decision-makers' motivation to be or appear consistent. For example, to avoid having inconsistencies revealed, decisions may become more favorable toward the side that is more likely to put a decision under scrutiny. As a context, we focus on the legal domain in which many high-stakes decisions are made in sequence and in which there are clear institutional processes of decision scrutiny. We analyze two field data sets: 386,109 U.S. immigration judges' decisions on asylum requests and 20,796 jury decisions in 18th century London criminal court. We distinguish between five mechanisms that can drive serial position effects and examine their predictions in these settings. We find that consistent with motivation-based explanations of serial position effects, but inconsistent with fatigue-based explanations, decisions become more lenient as a function of serial position, and the effect persists over breaks. We further find, as is predicted by motivational accounts, that the leniency effect is stronger among more experienced decision-makers. By elucidating the different drivers of serial position effects, our investigation clarifies why they are common, when they are expected, and how to reduce them. (PsycInfo Database Record (c) 2022 APA, all rights reserved).
... Indeed, the maintenance of forecast consistency is considered important by many (Perry and Green 1982;Quarantelli 1984;Drabek 1999), including the National Oceanic and Atmospheric Administration (NOAA 2016). Moreover, evidence from outside of the weather domain suggests that consumers believe that consistency between two estimates from the same source is a signal of skill and should be maintained when reputation is at stake (Falk and Zimmermann 2017). In addition, people can detect trends in inconsistent forecasts that influence their expectations about future forecasts. ...
... This may be because the negative impact of inaccuracy on trust is so powerful that it overwhelms the impact of inconsistency. Indeed, if consistency effects trust because it is regarded as a signal of skill, as previous work has suggested (Falk and Zimmermann 2017), inaccuracy may negate that impression. Taken together, these results suggest that any gain in trust from consistency may well be lost if the forecast turns out to be inaccurate. ...
... Here too, in all three experiments, consistency had only a small effect, smaller than the effect on postoutcome trust and much smaller than the effect on trust due to inaccuracy. This contradicts the intuition that the diagnostic relevance of consistency (Falk and Zimmermann 2017) should be greater in the absence of accuracy information. One possible explanation for the smaller impact of inconsistency pre-than postoutcome, is that in the inconsistent forecast pairs, when the second forecast was accurate (by the definition used here) the first forecast was inaccurate. ...
For high-impact weather events, forecasts often start days in advance. Forecasters believe that consistency among subsequent forecasts is important to user trust and can be reluctant to make changes when newer, potentially more accurate information becomes available. However, to date, there is little empirical evidence for an effect of inconsistency among weather forecasts on user trust, although the reduction in trust due to inaccuracy is well documented. The experimental studies reported here compared the effects of forecast inconsistency and inaccuracy on user trust. Participants made several school closure decisions based on snow accumulation forecasts for one and two days prior to the target event. Consistency and accuracy were varied systematically. Although inconsistency reduced user trust, the effect of the reduction due to inaccuracy was greater in most cases suggesting that it is inadvisable for forecasters to sacrifice accuracy in favor of consistency.
... In contrast, algorithms with a stronger tendency to exploit are by design more consistent over time in their choices. In settings with sequential choices, consistency, as opposed to changing behavior or choices, could be perceived as a signal of expertise (e.g., Falk and Zimmermann, 2017;Fehrler and Hughes, 2018;Soll et al., 2022), which would make exploitative algorithms appear as having a higher ability. Indeed, Ihssen et al. (2016) found that in a reversal learning task, participants were more willing to follow choices that were consistent over time than choices that varied between options. ...
... Even though the quality of the advice, i.e., the algorithm's success rate, was equal, participants were only willing to frequently accept the advice when the algorithm was less explorative and thus gave more consistent advice over time. One reason could be that consistency is regarded as a signal of expertise (Falk and Zimmermann, 2017;Fehrler and Hughes, 2018), which is supported by participants' reports of a higher usefulness of the algorithm's advice in the exploitative and balanced treatments. Following this line of reasoning, participants may have hesitated to follow recommendations to explore because they interpreted exploration as randomness or a demonstrated lack of expertise. ...
... While worker average performance is important, other statistics of the service time distribution also contribute to performance measurement. In many settings, quantiles of service times are used as measures of service quality (Hong 2009), and consistency (low variance) can be a signal of competency (Falk and Zimmermann 2016). In addition, queueing theory shows that service time variability contributes to wait times and queue lengths (Kingman 1961). ...
... where performance targets are defined by service time quantiles (Hong 2009), where performance consistency is important (Falk and Zimmermann 2016), or where queue length is partly driven by service time variability (Kingman 1961). ...
... Each row of the choice lists consists of a choice between two two-outcome lotteries. Inspired by Eeckhoudt and Schlesinger (2006) and Ebert and Wiesen (2014) , we chose one lottery to be a mean-preserving spread of the other, with a sure payoff (a risk premium) being added or subtracted. A noteworthy feature of this method is that it allows quantifying subjects' risk attitudes without assuming a specific utility function. ...
... Following these criteria, we chose a measure using two-outcome lotteries and mean-preserving spreads of these lotteries. Our method was inspired by Ebert and Wiesen (2011 ; ) whose experimental measures are based on the model-independent concept of risk apportionment ( Eeckhoudt and Schlesinger, 2006 ). Ebert and Wiesen (2014) classify an individual as risk-averse if she prefers a lottery L = (c L, 1 , p L, 1 ; c L, 2 , p L, 2 ) = (x − r, 50% ; x − k, 50%) over the lottery 6 A within-subject design would have had the advantage of providing us with a baseline measure of risk attitudes at the individual instead of group level. ...
A core prediction of recent “dual-self” models is that risk attitudes depend on self-control. While these models have received a lot of attention, empirical evidence regarding their predictions is lacking. We derive hypotheses from three prominent models for choices between risky monetary payoffs under regular and reduced self-control. We test the hypotheses in a lab experiment, using a well-established ego depletion task to reduce self-control, and measuring risk attitudes via finely graduated choice lists. Manipulation checks document the effectiveness of the depletion task. We find no systematic evidence in favor of the theoretical predictions. In particular, depletion does not increase risk aversion.
... We use the same question for the anchor and the valuation task as this is the way it is typically operationalised in the anchoring literature (Ariely et al., 2003;Fudenberg et al., 2012;Maniadis et al., 2014;Sugden et al., 2013). One interpretation of anchoring effects observed with this experimental design could be that anchoring is partially the result of preferences for consistency (Eyster, 2002;Falk & Zimmermann, 2017;Yariv, 2002). If subjects state their willingness to sell the lottery for a price of £ , then preferences for consistency imply that their minimum WTA for the same lottery should not be higher than £ . ...
... To minimize spillovers between the experimental and the survey measures, for example, because individuals might try to avoid cognitive dissonance (Festinger 1957) and strive for giving consistent responses (Falk and Zimmermann 2016;, we never ran survey and experiment for the same preference during the same session. More specifically, we conducted all experiments Falk et al.: Preference Survey Module Management Science, Articles in Advance, pp. ...
Incentivized choice experiments are a key approach to measuring preferences in economics but are also costly. Survey measures are a low-cost alternative but can suffer from additional forms of measurement error due to their hypothetical nature. This paper seeks to leverage the strengths of both approaches by proposing a new survey module on risk aversion, time discounting, trust, altruism, positive and negative reciprocity, in which survey items are selected based on ability to predict choices in corresponding, incentivized experiments. The methodology and results provided in the paper can also potentially provide a model for researchers who have specific requirements and want to design their own modules.
This paper was accepted by Yan Chen, behavioral economics and decision analysis.
... For instance, a left-wing incumbent may obtain better results than her right-wing opponent in a crisis context where social measures are needed; The reverse may hold in a context of economic expansion where laisser-faire can favor firms innovation. As in Falk and Zimmermann (2017) or Tajika (2021), policy consistency can also signal higher ability: The incumbent may thus inefficiently persist (Ashworth et al. 2010b) in failing policies. Some moves on the international scene can be more likely to succeed, in a complex geopolitical context, if the incumbent has political support the advantage that arises from the existence of a larger proportion of partisans for the incumbent. ...
A reelection-seeking politician makes a policy decision that can reveal her private information. This information bears on whether her political orientation and capabilities will be a good fit to future circumstances. We study how she may choose inappropriate policies to hide her information, even in the absence of specific conflicts of interests, and how voters’ conformism affects her incentives to do so. Conformism is independent from policies and from voters’ perceptions. Yet we identify a ‘conformism advantage’ for the incumbent that exists only when there is also an incumbency advantage. Conformism changes the incentives of the incumbent and favors the emergence of an efficient, separating equilibrium. It may even eliminate the pooling equilibrium (that can consist in inefficient persistence). Conformism has a mixed impact on social welfare however: it improves policy choices and the information available to independent voters, but fosters inefficient reelection in the face of a stronger opponent. When the incumbent is ‘altruistic’ and values social welfare even when not in power, she partly internalizes this latter effect. The impact of conformism is then non monotonous.
... Finally, Falk and Zimmermann (2017) documents that respondents may be driven by preference for consistent answers. Whereas we cannot rule out the possibility that the answer to the first elicitation question anchored participants' answer to the second elicitation question, a preference for consistency would imply a stickiness of political preferences leading to no preference changes (i.e., going against our hypothesis). ...
Voters often express support for a candidate whose policy platforms differ from their ideal policy preferences. We argue that under these circumstance acts of expressing support can causally
change voters’ policy preferences. We conceptualize our arguments in a theoretical model of policy preference changes rooted in cognitive dissonance theory. A pre-registered, online experiment with 1,200 U.S. participants confirms our main hypotheses. As predicted by cognitive
dissonance theory, voters align their policy preferences with those of the supported candidate. The more important the policy issue, the sharper the change in preferences. We also find that larger pre-support policy distance and higher effort in expressing support increases the magnitude of preference changes. Our results suggest that policy preferences can change mechanically after voters express support for a candidate.
... Reciprocal connectivity requires participants to respond to each other in their contributions, for example by taking up other participants' points of view (e.g., Graham & Witschge, 2003;Stromer-Galley, 2007). The expectation of consistency requires communication participants not to change their minds erratically, as consistency signals reliability and predictability (Falk & Zimmermann, 2017). ...
Research on incivility in political communication usually defines uncivil communication as a violation of established norms. Few studies, however, have specified these norms and corroborated them using relevant theoretical concepts. This article aims at strengthening the foundations of incivility research by analytically reconstructing the potential normative expectations of communication participants toward the behavior of others in offline and online political communication. We propose that these expectations can be considered as communication norms, which enable cooperative communication in political debates and conflicts. We use action theory, evolutionary anthropology, and linguistics to propose a norm concept that differentiates five communication norms: an information norm, a modality norm, a process norm, a relation norm, and a context norm. Drawing on these norms, we propose new definitions of incivility and civility. We also provide a comprehensive typology of norm violations that can be used as a heuristic for empirical research.
... Third, our results have practical implications since variability is a driver of system performance in most service systems (e.g., through queue lengths) and manufacturing systems (e.g., through quality control). Similarly, recent work has found that, through principles of behavioral economics, performance consistency can be a credible signal of competency (Falk and Zimmermann 2016). Therefore, our results suggest that the benefits of the learning curve may have been underestimated for practical purposes. ...
The expansive learning curve literature in operations management has established how various facets of prior experience improve average performance. In this paper, we explore how increased cumulative experience affects performance variability or consistency. We use a two-stage estimation method of a heteroskedastic learning curve model to examine the relationship between experience and performance variability among paramedics at the London Ambulance Service. We find that, for paramedics with lower experience, an increase in experience of 500 jobs reduces the variance of task completion time by 8.7%, in addition to improving average completion times by 2.7%. Similar to prior results on the average learning curve, we find a diminishing impact of additional experience on the variance learning curve. We provide an evidence base for how to model the learning benefits of cumulative experience on performance in service systems. Our findings imply that the benefits of learning are substantially underestimated if the consistency effect is ignored. Specifically, our estimates indicate that queue lengths (or wait times) might be overestimated by as much as 4% by ignoring the impact of the variance learning curve in service systems. Furthermore, our results suggest that previously established drivers of productivity should be revisited to examine how they affect consistency, in addition to average performance.
This paper was accepted by Charles Corbett, operations management.
... Third, our results have practical implications since variability is a driver of system performance in most service systems (e.g., through queue lengths) and manufacturing systems (e.g., through quality control). Similarly, recent work has found that, through principles of behavioral economics, performance consistency can be a credible signal of competency (Falk and Zimmermann 2016). Therefore, our results suggest that the benefits of the learning curve may have been underestimated for practical purposes. ...
jats:p> The expansive learning curve literature in operations management has established how various facets of prior experience improve average performance. In this paper, we explore how increased cumulative experience affects performance variability or consistency. We use a two-stage estimation method of a heteroskedastic learning curve model to examine the relationship between experience and performance variability among paramedics at the London Ambulance Service. We find that, for paramedics with lower experience, an increase in experience of 500 jobs reduces the variance of task completion time by 8.7%, in addition to improving average completion times by 2.7%. Similar to prior results on the average learning curve, we find a diminishing impact of additional experience on the variance learning curve. We provide an evidence base for how to model the learning benefits of cumulative experience on performance in service systems. Our findings imply that the benefits of learning are substantially underestimated if the consistency effect is ignored. Specifically, our estimates indicate that queue lengths (or wait times) might be overestimated by as much as 4% by ignoring the impact of the variance learning curve in service systems. Furthermore, our results suggest that previously established drivers of productivity should be revisited to examine how they affect consistency, in addition to average performance. This paper was accepted by Charles Corbett, operations management. </jats:p
... From a real options perspective, their lower hiring cost compared to poaching stars from other established firms might be attractive because of their high upside potential, despite the risk of bringing in bad hires. Ultimately, this study cautions against investigating the effect of an entrepreneurial spell on future wages without taking into account previous signals of ability, as employers do not consider the signal of an entrepreneurial experience in isolation but in combination with the signals emerging from the past paid employment experience (Falk & Zimmermann, 2017). ...
Research Summary
This article advances a theory to explain why a spell of entrepreneurship affects the future wages of entrepreneurs returning to the wage sector. We propose that entrepreneurship holds a low rather than a negative information value, increasing the uncertainty around a job applicant's future productivity. Employers respond to this uncertainty by discounting the offered wage. The theory predicts that uncertainty in hiring—and thus the wage penalty—is more pronounced for entrepreneurs (a) who were in the upper tail of the wage distribution before the entrepreneurial spell, (b) who exited entrepreneurship quickly, and (c) who are hired by small employers. We test and find empirical support for these predictions using a novel dataset of matched entrepreneurs and employees from Belgium.
Managerial Summary
We investigate the effect of past entrepreneurial experience on the future wages of entrepreneurs who go back to paid employment. We propose that former entrepreneurs receive a pay cut because employers consider them as risky hires. In line with our theory, we find that former entrepreneurs are penalized the most (a) if they were highly paid employees before becoming founders, (b) if they were entrepreneurs only for a few years, (c) and if they are hired by small employers. Our findings caution star employees from experimenting with an entrepreneurial career, as failing fast is costly. Moreover, we suggest that firms value probationary contracts to lower the risk of hiring entrepreneurs while learning about their skills.
... Although there is less research on this topic, it has been proposed that people prefer consistency, both from themselves and from others (Tedeschi, Schlenker, & Bonoma, 1971). In line with this, people infer that a person who gives more consistent estimates of an uncertain quantity over time is more skilled than someone who changes his/her estimate (Falk & Zimmermann, 2017), and believe politicians should not change their mind -unless of course they end up agreeing with the participant's own viewpoint (Croco, 2016). ...
Agreements and disagreements between expert statements influence lay people's beliefs. But few studies have examined what is perceived as a disagreement. We report six experiments where people rated agreement between pairs of probabilistic statements about environmental events, attributed to two different experts or to the same expert at two different points in time. The statements differed in frame, by focusing on complementary outcomes (45% probability that smog will have negative health effects vs. 55% probability that it will not have such effects), in probability level (45% vs. 55% probability of negative effects), or in both respects. Opposite frames strengthened disagreement when combined with different probability levels. Approximate probabilities can be “framed” in yet another way by indicating reference values they are “over” or “under”. Statements that use different directional verbal terms (over vs. under 50%) indicated greater disagreement than statements with the same directional term but different probability levels (over 50% vs. over 70%). Framing and directional terms similarly affected consistency judgments when both statements were issued by the same expert at different occasions. The effect of framing on perceived agreement was significant for medium (10 and 20 percentage points) differences between probabilities, whereas the effect of directional term was stable for numerical differences up to 40 percentage points. To emphasize agreement between different estimates, they should be framed in the same way. To accentuate disagreements or changes of opinion, opposite framings should be used.
... The scale develops from earlier works byBudner (1962),Rydell andRosen (1966), andMacDonald Jr. (1970).11 For example, individuals aim to be consistent in order to avoid cognitive dissonance(Festinger 1957;Falk and Zimmermann 2017). ...
Ambiguity preferences are important to explain human decision-making in many areas in economics and finance. To measure individual ambiguity preferences, the experimental economics literature advocates using incentivized laboratory experiments. Yet, laboratory experiments are costly, time-consuming and require substantial administrative effort. This study develops an experimentally validated ambiguity preference survey module that can reliably measure ambiguity preferences when carrying out laboratory experiments is impractical. This toolkit may have wide applications, including end-of-session lab questionnaires, large scale surveys and financial client assessments.
... Will such revisions also affect perceptions of certainty? One possibility is that recipients will see any kind of outcome revision as indicating uncertainty, since a change, in a sense, indicates that the first forecast was wrong or that the forecaster has changed his mind, and people may associate inconsistency in estimates with low skills (Falk & Zimmermann, 2017). However, forecasts revised in an upward direction, such as IPCC's change of the climate sensitivity estimate from 1.5-4.5°C to 2.0-4.5°C, may be seen as more certain, for several reasons. ...
What do lay people think about revised forecasts of future outcomes? A series of experiments
show that when climate change forecasts are revised in an upward direction (e.g., a higher sea
level rise is predicted in light of new information), the forecast is perceived as more certain, in
contrast to revisions in a downward direction. This association is bidirectional, so that people
also think a forecast that has become more certain (uncertain) indicates a stronger (weaker)
outcome. Furthermore, when the timing of the outcome is revised, a predicted sooner
occurrence (a “stronger” outcome) was judged to be more certain than a predicted delayed
occurrence. Upward revisions may also lead to more positive impressions of the forecast and
the forecaster, with clear implications for the communication of uncertainty, both for climate
change and in other domains. Different theoretical explanations for the results are discussed.
... Risky process In the Ex-Ante conditions, after making their decisions the decision makers were given a chance to reconsider their choice after learning about the actual outcome of the random process. We may expect that there are various psychological reasons not to change the decision ex-post, including a preference for consistency (e.g., Falk and Zimmermann 2016) or simple inertia (Samuelson and Zeckhauser 1988), even if the decision maker regrets her initial choice. Consistent with this view, we find only modest switching, with overall 21.7% of the decision makers reconsidering their initial choice; specifically, 13.0% change after an advantageous outcome obtained (N = 23), and 30.4% change after a disadvantageous outcome obtained (N = 23). ...
Literature on fairness preferences distinguishes between outcome fairness, concerning the final allocation of payoffs, and process fairness, concerning the expected allocation of payoffs. It is not obvious, however, whether process fairness can consistently be implemented. Once uncertainty is resolved and outcomes are determined, the ex-ante procedurally fair decision maker may become consequentialist ex-post, and reconsider her choice on the basis of the observed outcomes. We present experimental evidence on dynamic consistency of social preferences under both known risk and ambiguity. A significant share of people subscribe to process fairness both before and after the resolution of uncertainty.
Little is known about how people-centric factors affect the shape of service time distributions despite distributional statistics (variance or quantiles) being key drivers of system performance in many service industries. We investigate the impact of one people-centric factor—worker fatigue—on the average, variance, and quantiles of service times in paramedic operations. Our analysis uses data on the performance of 368,634 paramedic teams in the London Ambulance Service over 10 years. We measure fatigue by the number of prior jobs a paramedic crew has completed during a shift and estimate its impact on the time it takes the crew to respond to incidents and bring patients to hospitals. Using a recentered influence function regression approach with multiple fixed effects, we find that the average time to hospital increases by 5% throughout the course of an average shift. In addition, the workers become less consistent with fatigue; service time variance increases by 39% during a normal shift. Furthermore, we find that in addition to an upward shift in mean service times, both the upper and lower tails of the distribution have more weight for fatigued paramedics. These effects are driven mostly by the performance of paramedics at the scene, rather than their driving to or from the incident. The distributional effects of fatigue are only slightly mitigated by increased experience or reduced system workload. Our work demonstrates that the impact of people-centric factors can be highly nonuniform across the service time distribution.
This paper was accepted by Jayashankar Swaminathan, operations management.
Funding: The authors are grateful for financial research support from the Wisconsin Alumni Research Foundation at the University of Wisconsin–Madison.
Supplemental Material: The data files and e-companion are available at https://doi.org/10.1287/mnsc.2023.4855 .
When forecasts for a major weather event begin days in advance, updates may be more accurate but inconsistent with the original forecast. Evidence suggests that resulting inconsistency may reduce user trust. However, adding an uncertainty estimate to the forecast may attenuate any loss of trust due to forecast inconsistency, as has been shown with forecast inaccuracy. To evaluate this hypothesis, this experiment tested the impact on trust of adding probabilistic snow-accumulation forecasts to single-value forecasts in a series of original and revised forecast pairs (based on historical records) that varied in both consistency and accuracy. Participants rated their trust in the forecasts and used them to make school-closure decisions. One-half of the participants received single-value forecasts, and one-half also received the probability of 6 in. or more (decision threshold in the assigned task). As with previous research, forecast inaccuracy was detrimental to trust, although probabilistic forecasts attenuated the effect. Moreover, the inclusion of probabilistic forecasts allowed participants to make economically better decisions. Surprisingly, in this study inconsistency increased rather than decreased trust, perhaps because it alerted participants to uncertainty and led them to make more cautious decisions. Furthermore, the positive effect of inconsistency on trust was enhanced by the inclusion of probabilistic forecast. This work has important implications for practical settings, suggesting that both probabilistic forecasts and forecast inconsistency provide useful information to decision-makers. Therefore, members of the public may benefit from well-calibrated uncertainty estimates and newer, more reliable information.
Significance Statement
The purpose of this study was to clarify how explicit uncertainty information and forecast inconsistency impact trust and decision-making in the context of sequential forecasts from the same source. This is important because trust is critical for effective risk communication. In the absence of trust, people may not use available information and subsequently may put themselves and others at greater-than necessary risk. Our results suggest that updating forecasts when newer, more reliable information is available and providing reliable uncertainty estimates can support user trust and decision-making.
Beliefs are, in many ways, central to psychology and, in turn, consistency is central to belief. Theories in philosophy and psychology assume that beliefs must be consistent with each other for people to be rational. That people fail to hold fully consistent beliefs has, therefore, been the subject of much theorizing, with numerous mechanisms proposed to explain how inconsistency is possible. Despite the widespread assumption of consistency as a default, achieving a consistent set of beliefs is computationally intractable. We review research on consistency in philosophy and psychology and argue that it is consistency, not inconsistency, that requires explanation. We discuss evidence from the attitude, belief, and persuasion literatures, which suggests that accessibility of beliefs in memory is one possible mechanism for achieving a limited, but psychologically plausible, form of consistency. Finally, we conclude by suggesting future directions for research beginning from the assumption of inconsistency as the default.
This article is categorized under: Psychology > Reasoning and Decision Making
Psychology > Theory and Methods
Philosophy > Knowledge and Belief
To understand why naiveté about present-biased behavior is so prevalent and persistent, our experiment investigates how well participants learn from their past behavior in a real-effort task. While participants display naive present-biased behavior initially, our novel methodology allows us to establish that their updating is unbiased and would eliminate naiveté in the long run. Moreover, learning is unencumbered by a change in environment. Our results suggest that persistent naiveté does not result from a fundamental inferential bias. However, participants underestimate their future learning—a bias that may lead to underinvestment in experimentation and a failure to activate self-regulation mechanisms.
We use an experiment to test the hypothesis that adding discretionary adjustment to performance-based pay strengthens the sorting of employees based on how strongly they identify with the organization’s objectives. Our conceptualization of identification is grounded in identity economics, which predicts that employees who identify strongly with the organization’s objectives exert greater effort toward those objectives than employees who identify weakly with those objectives. Building on this conceptualization, we expect that employees anticipate that managers will adjust performance-based pay more (less) favorably when employees reveal strong (weak) identification with the organization’s objectives. Thus, when managers can adjust performance-based pay, performance-based pay contains a feature that benefits (disadvantages) employees with strong (weak) identification, which we expect to strengthen the sorting of employees based on their identification. Consistent with our hypothesis, we find that the difference in preferences for performance-based pay between employees with strong and weak identification is larger when discretionary adjustment accompanies performance-based pay than when it does not. Our results also confirm that employee identification increases employee costly effort exertion toward the organization’s objectives. We contribute to the management accounting literature on discretion in performance evaluation by documenting a previously undocumented benefit of discretionary adjustment.
We study the effect of entitlement on the willingness to lie. We set up a model of lying where individuals feel more or less entitled to their endowment depending on how they earned it. When given the opportunity to lie to keep their endowment, individuals who feel more entitled are encouraged to lie while others are discouraged. To test the model predictions we use a laboratory experiment where we compare the lying behavior of participants endowed with a high endowment and participants endowed with a low endowment. In one treatment, the allocation of the endowment is decided by participants’ performance, and in the other, it is determined by a random draw. Our study shows that deservingness influences lying in an intuitive direction: when participants performance determines income, those who earn less money lie less than those who earn more. We do not find differences in lying when participants perform the same task but lie to keep windfall endowments.
To avoid unfavorable inferences about her ability, an expert might cling to her original opinion and ignore valuable new information in formulating subsequent opinions. Conceivably, the expert might decline an initial opportunity to offer an opinion, delaying the opinion formation until more accurate information has arrived. However, we show that reputational concerns often lead an expert to express an opinion at the first opportunity, thereby making a snap decision.
We study the effect of entitlement on the willingness to lie. We set up a model of lying where individuals feel more or less entitled to their endowment depending on how they earned it. When given the opportunity to lie to keep their endowment, individuals who feel more entitled are encouraged to lie while others are discouraged. To test the model predictions we use a laboratory experiment where we compare the lying behavior of participants endowed with a high endowment and participants endowed with a low endowment. In one treatment, the allocation of the endowment is decided by participants’ performance, and in the other, it is determined by a random draw. Our study shows that deservingness influences lying in an intuitive direction: when participants performance determines income, those who earn less money lie less than those who earn more. We do not find differences in lying when participants perform the same task but lie to keep windfall endowments.
Experts and workers in many domains make multiple similar but independent decisions in sequence. Previous research has shown that the serial position of the case in the sequence, an irrelevant factor, often influences the decision. Yet, the conditions under which serial position effects emerge remain unclear. Explanations for these effects tend to focus on the role of decision makers’ fatigue, but these effects emerge also when fatigue is unlikely. Here, we highlight the importance of considering decision makers’ motivation to produce consistent sets of decisions. We focus on the legal domain in which many high-stakes decisions are made in sequence. We analyze two field datasets: 386,109 US immigration judges’ decisions on asylum requests and 20,796 jury decisions in 18th century London criminal court. We distinguish between five mechanisms that can drive serial position effects and examine their predictions for these settings. We find that consistent with motivation-based explanations of serial position effects, but inconsistent with fatigue-based explanations, in both settings decisions become more lenient as a function of serial position, an effect that persists over breaks and contrasts previous findings concerning sequential parole hearings. We further find the leniency effect is stronger among more experienced decision makers. To complement the analysis, we run controlled experiments with laypeople and find similar results. Theoretically, we suggest that decisions may become more favorable towards the side that is more likely to put a decision under scrutiny. In many contexts, including legal decisions, this process implies that it is often “best to be last”.
We show that a one-off incentive to bias advice has a persistent effect on advisers' own actions and their future recommendations. In an experiment, advisers obtained information about a set of three differently risky investment options to advise less informed clients. The riskiest option was designed such that it is only preferred by risk-seeking individuals. When advisers are offered a bonus for recommending this option, half of them recommend it. In contrast, in a control group without the bonus only four percent recommend it. After the bonus was removed, its effect remained: In a second recommendation for the same options but without a bonus, those advisers who had previously faced it are almost six times more likely to recommend the riskiest option compared to the control group. A similar increase is found when advisers make the same choice for themselves. To explain our results we provide a theory based on advisers trying to uphold a positive self-image of being incorruptible. Maintaining a positive self-image then forces them to be consistent in the advice they give, even if it is biased.
We investigate the credibility of nonbinding preplay statements about cooperative behavior, using data from a high-stakes TV game show in which contestants play a variant on the classic Prisoner’s Dilemma. We depart from the conventional binary approach of classifying statements as promises or not, and propose a more fine-grained two-by-two typology inspired by the idea that lying aversion leads defectors to prefer statements that are malleable to ex-post interpretation as truths. Our empirical analysis shows that statements that carry an element of conditionality or implicitness are associated with a lower likelihood of cooperation, and confirms that malleability is a good criterion for judging the credibility of cheap talk.
This paper was accepted by Elke Weber, judgment and decision making.
The dissertation consists of four chapters. The first two chapters are devoted to exploring information acquisition and disclosure in contests. The third chapter is devoted to exploring how risk attitude affects bidding behaviour in all-pay auctions. the last chapter is denoted to exploring behavioural biases in advice-giving. In Chapter 1, I study player's incentive to spy on opponents' private information in contests. I show that each player's equilibrium effort is non-decreasing (non-increasing) in the posterior probability that the opponent has the same (a different) valuation. Accounting for the cost of spying, players are strictly better off than not spying on each other at all. In Chapter 2, I focus on how a contest organiser should disclose information in order to achieve certain objectives. In particular, I compare private signals with public signals. I show that there is no general ranking of the two signals in terms of the performance of maximizing players' expected payoff, but public signals outperforms private signals in maximizing expected effort. In Chapter 3 (co-authors with David Ong and Ella Segev), we extend previous theoretical work on n-players complete information all-pay auction to incorporate heterogeneous risk and loss averse utility functions. We provide sufficient and necessary conditions for the existence of equlibria with a given set of active players with any strictly increasing utility functions and characterize the players' equilibrium mixed strategies. Finally, in Chapter 4 (co-authored with Tobias Gesche), we show experimental evidence that a one-ff incentive to bias advice has a persistent effect on advisers' own actions and their future recommendations.
We show that a one-off incentive to bias advice has a persistent effect on advisers’ own actions and their future recommendations. In an experiment, advisers obtained information about a set of three differently risky investment options to advise less informed clients. The riskiest option was designed such that it is only preferred by risk-seeking individuals. When advisers are offered a bonus for recommending this option, half of them recommend it. In contrast, in a control group without the bonus only four percent recommend it. After the bonus was removed, its effect remained: In a second recommendation for the same options but without a bonus, those advisers who had previously faced it are almost six times more likely to recommend the riskiest option compared to the control group. A similar increase is found when advisers make the same choice for themselves. To explain our results we provide a theory based on advisers trying to uphold a positive self-image of being incorruptible. Maintaining a positive self-image then forces them to be consistent in the advice they give,
even if it is biased.
A central assumption in economics is that people misreport their private information if this is to their material benefit. Several recent models depart from this assumption and posit that some people do not lie or at least do not lie maximally. These models invoke many different underlying motives including intrinsic lying costs, altruism, efficiency concerns, or conditional cooperation. To provide an empirically-validated microfoundation for these models, it is crucial to understand the relevance of the different potential motives. We measure the extent of lying costs among a representative sample of the German population by calling them at home. In our setup, participants have a clear monetary incentive to misreport, misreporting cannot be detected, reputational concerns are negligible and altruism, efficiency concerns or conditional cooperation cannot play a role. Yet, we find that aggregate reporting behavior is close to the expected truthful distribution suggesting that lying costs are large and widespread. Further lab experiments show that this result is not driven by the mode of communication.
In this paper we ask if reports of private information about skills, abilities or achievements are affected by image concerns. We develop a simple model that illustrates how image utility can lead to misreporting of private information in contexts where truthful reports maximize monetary outcomes. In addition, we test the model's predictions in a controlled lab experiment. In the experiment, all subjects go through a series of quiz questions and subsequently report a performance measure. We vary if reports are made to an audience or not and find evidence for image effects. In the audience treatment, stated reports are significantly higher than in the private treatment. This suggests that overconfident appearance might be a consequence of social approval seeking. We also find that men state higher self-assessments than women. This gender difference seems to be driven by men responding more strongly to the presence of an audience.
Examines various interpretations of cognitive dissonance. A review of the conditions which have reliably produced dissonance indicates that the phenomenon will occur whenever an individual can be held intentionally responsible for his behaviors and is concerned with the attributions others make about him. A social-psychologically based theory of impression management is offered to account for the results of dissonance studies. According to the theory, an individual learns that he must appear consistent in order to maintain his credibility as a social communicator. Additionally, he is constrained by normative considerations. Dissonance reduction will only be observed when a person believes that an outside o imputes intentionality to both of 2 contradictory statements or actions, and the 2nd action is believed to be counternormative or harmful in nature. The theory integrates the various viewpoints of L. Festinger, D. Bem, V. Allen, and M. Rosenberg, is supported by the existing dissonance literature, is rooted in socialization processes, and offers an alternative to the traditional intrapsychic explanations of dissonance. (50 ref.) (PsycINFO Database Record (c) 2012 APA, all rights reserved)
This paper explores the consequences of cognitive dissonance, coupled with time-inconsistent preferences, in an intertemporal decision problem with two distinct goals: acting decisively on early information (vision) and adjusting flexibly to late information (flexibility). The decision maker considered here is capable of manipulating information to serve her self-interests, but a tradeoff between distorted beliefs and distorted actions constrains the extent of information manipulation. Building on this tradeoff, the present model provides a unified framework to account for the conformity bias (excessive reliance on precedents) and the confirmatory bias (excessive attachment to initial perceptions).
Everyday observation reveals that both individuals and organizations often become overly committed to losing courses of action; in a sense, throwing good money after bad. More than 10 years of research on this escalation problem shows that persistence is associated with at least four major clases of determinants: project, psychological, social, and organizational variables. The influence of these four sets of variables evolves over time, forming a dynamic model of behavior in escalation situations.
z-Tree (Zurich Toolbox for Ready-made Economic Experiments) is a software for developing and conducting economic experiments. The software is stable and allows programming almost any kind of experiments in a short time. In this article, I present the guiding principles behind the software design, its features, and its limitations. Copyright Economic Science Association 2007
We develop a theory of prosocial behavior that combines heterogeneity in individual altruism and greed with concerns for social reputation or self-respect. Rewards or punishments (whether material or image-related) create doubt about the true motive for which good deeds are performed, and this ?overjustification effect? can induce a partial or even net crowding out of prosocial behavior by extrinsic incentives. We also identify the settings that are conducive to multiple social norms and, more generally, those that make individual actions complements or substitutes, which we show depends on whether stigma or honor is (endogenously) the dominant reputational concern. Finally, we analyze the socially optimal level of incentives and how monopolistic or competitive sponsors depart from it. Sponsor competition is shown to potentially reduce social welfare. (JEL D11, D64, D82, Z13)
We provide evidence that long-term relationships between trading parties emerge endogenously in the absence of third party enforcement of contracts and are associated with a fundamental change in the nature of market interactions. Without third party enforcement, the vast majority of trades are initiated with private offers and the parties share the gains from trade equally. Low effort or bad quality is penalized by the termination of the relationship, wielding a powerful effect on contract enforcement. Successful long-term relations exhibit generous rent sharing and high effort (quality) from the very beginning of the relationship. In the absence of third-party enforcement, markets resemble a collection of bilateral trading islands rather than a competitive market. If contracts are third party enforceable, rent sharing and long-term relations are absent and the vast majority of trades are initiated with public offers. Most trades take place in one-shot transactions and the contracting parties are indifferent with regard to the identity of their trading partner. Copyright The Econometric Society 2004.
Beliefs are often found to be sticky and rather immune to new information. In this paper we highlight a specific mechanism that raises resistance to incorporate new information. We provide causal evidence that commitment to a first opinion leads to a neglect of new and challenging information. Being asked to write down an initial estimate before additional information is obtained significantly and substantially reduces the accuracy of final beliefs. Investigating the sources of this effect, we show that our findings are well explained by an internal desire to act consistently.
This article studies how a taste for consistency affects decision making. Our application is response behavior in surveys.
In particular, we show that the inclusion of questions can affect answers to subsequent related questions. The reason is that
participants want to respond in a consistent way. Studying three different surveys, we find a systematic effect of the inclusion
of additional questions. The effects are large and reveal how easy survey responses can be manipulated. For example, we find
that a subtle manipulation reduced the number of people who agreed that a murderer should be imprisoned for the rest of his
life by more than 20 percentage points. (JEL codes: C83, C91, D03)
We examine factors that may contribute to 'overconfidence' in relative ability on an intelligence test. We test experimentally for evidence of self-esteem concerns and instrumental strategic concerns. Errors in Bayesian updating are rare when the information does not involve own relative ability, but far more common when it does, suggesting self-esteem issues. There is also strong evidence that males state higher levels of confidence in relative ability when this precedes a tournament; as entry is predicted by relative confidence, this can be an effective deterrent. Inflating confidence can be part of an equilibrium strategy, providing a rationale for strategic overconfidence.
People dislike knowing that they would be better off having made different past choices. This paper models the behavior of someone with a taste for rationalizing past actions by taking current actions for which those past actions were optimal. When past and present actions are strategic complements, then a past action higher than optimal leads to a present action also higher than optimal-the sunk-cost effect-and a past action lower than optimal leads to a present action also lower than optimal-an "unsunk-cost effect. " A taste for consistency gives rise to the declining-price anomaly in sequential auctions. A rational monopolist may best respond to consumers who rationalize the past by limiting the menu of tariffs it offers them. People who underestimate their future taste for consistency overbid in wars of attrition and procrastinate in search models.
Strategic self presentations can have a far-reaching impact on an actor's identity. Subjects who presented themselves as sociable to an interviewer, compared with those who did not present themselves, later raised their self-appraisals of their own sociability, behaved more sociably (i.e., spoke sooner, more frequently, and longer) in a different situation, were viewed as more sociable by a confederate and by judges, and recalled personal experiences that indicated they were more sociable. Strategic self presentations thus produced both a phenomenological and a behavioral carry-over that influenced the actor's identity in a new situation with a new audience. Two further experiments explored the processes responsible for these effects and found that private self-reflection was not sufficient to produce the changes. Rather; public commitment to the identity portrayed in the self-presentation was a crucial antecedent of changes in self-appraisals.
An attempt is made to analyze interactive social behavior in terms of communicative acts. The basic assumption is, "that communication among humans performs the essential function of enabling two or more individuals to maintain simultaneous orientation toward one another as communicators and toward objects of communication." A rationale for this assumption is presented and then a set of propositions derived from and consistent with it are shown to be supported by much existing empirical evidence.
This paper examines individual decision making when decisions reflect on people's ability to learn. The authors address this problem in the context of a manager making investment decisions on a project over time. They show that, in an effort to appear as a fast learner, the manager will exaggerate his own information but ultimately he becomes too conservative, being unwilling to change his investments on the basis of new information. The authors' results arise purely from learning about competence rather than concavity or convexity of the rewards functions. They relate their results to the existing psychology literature concerning cognitive dissonance reduction. Copyright 1996 by University of Chicago Press.
This paper analyzes a model in which individuals care about both consumption (intrinsic utility) and social status. Status depends on public perceptions about an individual's predispositions rather than on the individual's actions. However, since predispositions are unobservable, actions signal predispositions and therefore affect status. When status is sufficiently important relative to intrinsic utility, many individuals conform to a rigid standard of behavior, despite heterogeneous intrinsic preferences. When status is relatively unimportant, no conformity emerges. The model produces both customs and fads, and it suggests an explanation for the development of multiple subcultures with distinct norms. Copyright 1994 by University of Chicago Press.
Cognitive dissonance theory predicts that the act of voting for a candidate leads to a more favorable opinion of the candidate in the future. We find support for the empirical relevance of cognitive dissonance to political attitudes. We examine the presidential opinion ratings of voting-age eligibles and ineligibles two years after the president's election. We find that eligibles show two to three times greater polarization of opinions than comparable ineligibles. We find smaller effects when we compare polarization in opinions of senators elected during high turnout presidential campaign years with senators elected during nonpresidential campaign years. (JEL D72)
In this paper we present a new design which allows us to draw inferences on the distribution of lying behavior among the population. Participants received a dice in order to determine their payoff anonymously. Whatever they reported to have rolled, they received as payoff. 39% of the subjects were honest and maximally 22% of them were lying completely. Interestingly we found subjects who lied but who did not maximize their income by doing so. Using additional experiments, we can show that a compelling explanation for this behavior is the desire to maintain a favorable self-concept, including honesty and non-greediness.
This paper experimentally examines image motivation - the desire to be liked and wellregarded by others - as a driver in prosocial behavior (doing good), and asks whether extrinsic monetary incentives (doing well) have a detrimental effect on prosocial behavior due to crowding out of image motivation. By definition, image depends on one’s behavior being visible to other people. Using this unique property we show that image is indeed an important part of the motivation to behave prosocially. Moreover, we show that extrinsic incentives interact with image motivation and are therefore less effective in public than in private. Together, these results imply that image motivation is crowded out by monetary incentives; which in turn means that monetary incentives are more likely to be counterproductive for public prosocial activities than for private ones.
Desire for social esteem is a source of prosocial behavior. We develop a model in which actors' utility of esteem depends on the audience. In a principal-agent setting, we show that the model can account for motivational crowding out. Control systems and pecuniary incentives erode morale by signaling to the agent that the principal is not worth impressing. The model also offers an explanation for why agents are motivated by unconditionally high pay and by mission-oriented principals. (JEL D01, D82)
We examine experimentally the impact of communication on trust and cooperation. Our design admits observation of promises, lies, and beliefs. The evidence is consistent with people striving to live up to others' expectations so as to avoid guilt, as can be modeled using psychological game theory. When players exhibit such guilt aversion, communication may influence motivation and behavior by influencing beliefs about beliefs. Promises may enhance trustworthy behavior, which is what we observe. We argue that guilt aversion may be relevant for understanding strategic interaction in a variety of settings, and that it may shed light on the role of language, discussions, agreements, and social norms in these contexts. Copyright The Econometric Society 2006.
2005) I’ll see it when I believe it-A simple model of cognitive consistency. Cowles Foundation Discussion Paper 1352
- L Yariv