George Loewenstein, Cass R. Sunstein, and Russell Golman, “Disclosure: Psychology Changes Everything.” Annual Review of Economics 6: 391–419, 2014 (pdf).
• A good deal of information is disclosed through government mandate, and much of this information would not be disclosed in the absence of the mandate. The information would not be disclosed perhaps because it would not benefit sellers, or because information has a public good aspect that lowers the incentive for any single private entity to produce and disseminate the information.
• Mandated disclosures involve some subtle costs, such as the time they take for consumers to read them, the subsequent loss of attention to other pieces of information, and even the emotional costs associated with graphic warnings, for instance.
• Mandated disclosures tend to occur when there are significant gaps in the information known to sellers and that known to buyers, and when the informational disadvantage threatens the interests of consumers. Disclosure also can be used to help consumers overcome their own departures from rational decision making: perhaps “behavioral market failures” provide a rationale for policies to limit internalities.
• Some information – such as a physician’s assertion that a certain treatment is needed – is not verifiable, and hence problems connected with this information cannot be solved simply through disclosure. But physicians might be required to disclose their interests (such as receiving royalties from the recommended treatment) if those interests are not fully aligned with patient interests.
• Disclosure of conflicting incentives does not fix every problem. The disclosing agent might view the disclosure as allowing for carte blanche, for any sort of self-interested advice. The recipient (principal) might feel compelled to follow the advice, to avoid the inference that the advice giver is viewed as untrustworthy.
• The technology of disclosure – who makes it, when, and what effort is made to render it noticeable – helps to determine its impact.
• Sellers have little reason to put effort into those dimensions of a good that potential consumers do not pay attention to. A producer of a less deadly cigarette might not want to disclose its relative safety, because to do so might make the fatal consequences of smoking more salient.
• Warning labels don’t seem to accomplish much; more generally, see Omri Ben-Shahar and Carl E. Schneider, More Than You Wanted to Know: The Failure of Mandated Disclosure, Princeton U. P., 2014.
• The absence of information should best be met, perhaps, by assuming the worst, as otherwise the information would have been provided. But people often do not draw this inference, even when it is rational to do so. This presents a potential rationale for mandating disclosure.
• Sometimes we want to be ignorant, sometimes information can lower our utility – a tendency that motivated the May, 2015 on-air radio killing of a young rabbit in Denmark.
• The tell-tale heart effect: mandated disclosure might cause producers to up their game, even if no consumers pay attention. Revelation of calorie counts might lead to lower calorie offerings, even if consumers do not respond to the calorie information. Maybe producers suffer from a spotlight effect, a belief that people are observing their disclosures more closely than really is the case.
• Simplified information, like restaurant health grades, is often more valuable to consumers than is more finely grained information.
• Comparative information – how does my energy use stack up against my neighbors? – might be more influential on energy usage than other types of usage disclosures. But the potential for perverse outcomes exists, too.
• Personal policies for information disclosure or non-disclosure on social media, for example, do not seem to be fully rational.
Friday, November 6, 2015
Educating Behaviorally with Koch, et al. (2015)
Alexander Koch, Julia Nafziger, and Helena Skyt Nielsen, “Behavioral Economics of Education.” Journal of Economic Behavior & Organization 115: 3-17, July 2015.
• Education decisions typically involve current investments of time and effort for (probabilistic) rewards that often will be realized only many years later. Education is rife with possibilities for less-than-rational decisions, including those resulting from low willpower. There is evidence that many educational decisions indeed fall prey to irrationality, such as decisions to drop out of high school.
• Soft skills, such as personality traits and willpower, are as (or more) important as cognitive skills in educational success.
• Family background and inputs, peer groups, school resources, and teacher skills, have large effects on scholastic success.
• Some teachers seem to help students build cognitive skills, and some aid non-cognitive skills. Nevertheless, “it is largely unknown what constitutes a good teacher [p. 5].”
• On average, it seems that girls get better grades and boys score higher on standardized tests. These outcomes seem to be partly explained by gender stereotypes and approaches to competition (which themselves might depend on gender stereotypes).
• Good grades require self-control skills. People who under-study perhaps overestimate their sophistication, or lack precommitment opportunities. Dropping out and disciplinary problems might be signs of poor self-control.
• Personal goal setting can help, by altering reference points and thereby leveraging loss aversion. Goals should not be so ambitious that they lead to despair and abandonment, however.
• Extra incentives (such as monetary bonuses) are generally good at motivating better performance with respect to attendance and enrollment.
• Incentives are tricky, as they work differently for different populations. For instance, some students perform better with an absolute grading scale, while others thrive in relative grading systems with only a few broad ranks.
• Effort and ability tend to be complements, so students do better if they have a positive view of their ability. A system that offers few high grades can induce students to not try, perhaps as a protective method against possibly learning negative information about their ability. A high reward offered to a child in exchange for a high grade makes it appear that the student is not expected to do well, which can lower the student’s self-assessment of her ability and subsequently undermine effort.
• Bad schools might need to pay bonuses to attract good teachers, but good schools, which offer more intrinsic rewards to teaching, can attract good teachers without extraordinary bonuses.
• In terms of encouraging one’s own effort and self-control, it is good to have a peer who has slightly less self-control. Then, that person’s victories can be encouraging, while that person’s failures can be distinguished from your own prospects by their relative lack of self-control.
• Education decisions typically involve current investments of time and effort for (probabilistic) rewards that often will be realized only many years later. Education is rife with possibilities for less-than-rational decisions, including those resulting from low willpower. There is evidence that many educational decisions indeed fall prey to irrationality, such as decisions to drop out of high school.
• Soft skills, such as personality traits and willpower, are as (or more) important as cognitive skills in educational success.
• Family background and inputs, peer groups, school resources, and teacher skills, have large effects on scholastic success.
• Some teachers seem to help students build cognitive skills, and some aid non-cognitive skills. Nevertheless, “it is largely unknown what constitutes a good teacher [p. 5].”
• On average, it seems that girls get better grades and boys score higher on standardized tests. These outcomes seem to be partly explained by gender stereotypes and approaches to competition (which themselves might depend on gender stereotypes).
• Good grades require self-control skills. People who under-study perhaps overestimate their sophistication, or lack precommitment opportunities. Dropping out and disciplinary problems might be signs of poor self-control.
• Personal goal setting can help, by altering reference points and thereby leveraging loss aversion. Goals should not be so ambitious that they lead to despair and abandonment, however.
• Extra incentives (such as monetary bonuses) are generally good at motivating better performance with respect to attendance and enrollment.
• Incentives are tricky, as they work differently for different populations. For instance, some students perform better with an absolute grading scale, while others thrive in relative grading systems with only a few broad ranks.
• Effort and ability tend to be complements, so students do better if they have a positive view of their ability. A system that offers few high grades can induce students to not try, perhaps as a protective method against possibly learning negative information about their ability. A high reward offered to a child in exchange for a high grade makes it appear that the student is not expected to do well, which can lower the student’s self-assessment of her ability and subsequently undermine effort.
• Bad schools might need to pay bonuses to attract good teachers, but good schools, which offer more intrinsic rewards to teaching, can attract good teachers without extraordinary bonuses.
• In terms of encouraging one’s own effort and self-control, it is good to have a peer who has slightly less self-control. Then, that person’s victories can be encouraging, while that person’s failures can be distinguished from your own prospects by their relative lack of self-control.
Sunday, November 1, 2015
Kahneman (2011), “Experienced Well-Being”
Daniel Kahneman, “Experienced Well-Being.” Chapter 37, pages 391-397, in Thinking, Fast and Slow, New York: Farrar, Straus and Giroux, 2011.
• “Happiness” conflates experienced utility with remembered utility. Questions about subjective well-being probably do not map well with experienced utility. In a previous chapter (chapter 35, pages 377-385), Kahneman indicates that people’s memories of a past event tend to neglect the duration of the event (which is quite important as the event is being experienced), and that their assessment of a past event primarily will reflect the average of the peak level (of pain, say, for an aversive episode) and the state at the end of the event.
• How to measure experienced, moment-by-moment, utility? One method (“experience sampling”) is to interrupt people at random times throughout the day, and to ask them what they are doing and how they are feeling about it. The Day Reconstruction Method provides another measure, where people are asked to map in detail their previous day’s activities, and their emotional state at the time of the activities.
• The U-index measures the proportion of time people spend in an unpleasant state. The U-index exhibits significant variance: “a small fraction of the population does most of the suffering [p. 394].”
• You probably can improve your personal U-index by switching your leisure pursuits away from passive forms (watching television) to active forms (socializing).
• Policy might be able to reduce the overall U-index, perhaps by encouraging socializing for elderly people or by reducing commuting times. A small decrease in the percentage of time spent in an unpleasant state is quite significant in terms of suffering averted.
• Poverty can bring unhappiness, but money does not add to experienced well-being for household incomes above $75,000.
• “Happiness” conflates experienced utility with remembered utility. Questions about subjective well-being probably do not map well with experienced utility. In a previous chapter (chapter 35, pages 377-385), Kahneman indicates that people’s memories of a past event tend to neglect the duration of the event (which is quite important as the event is being experienced), and that their assessment of a past event primarily will reflect the average of the peak level (of pain, say, for an aversive episode) and the state at the end of the event.
• How to measure experienced, moment-by-moment, utility? One method (“experience sampling”) is to interrupt people at random times throughout the day, and to ask them what they are doing and how they are feeling about it. The Day Reconstruction Method provides another measure, where people are asked to map in detail their previous day’s activities, and their emotional state at the time of the activities.
• The U-index measures the proportion of time people spend in an unpleasant state. The U-index exhibits significant variance: “a small fraction of the population does most of the suffering [p. 394].”
• You probably can improve your personal U-index by switching your leisure pursuits away from passive forms (watching television) to active forms (socializing).
• Policy might be able to reduce the overall U-index, perhaps by encouraging socializing for elderly people or by reducing commuting times. A small decrease in the percentage of time spent in an unpleasant state is quite significant in terms of suffering averted.
• Poverty can bring unhappiness, but money does not add to experienced well-being for household incomes above $75,000.
Stevenson and Wolfers (2008) on the Easterlin Paradox
Betsey Stevenson and Justin Wolfers, “Economic Growth and Subjective
Well-Being: Reassessing the Easterlin Paradox.” Brookings Papers on Economic Activity, pages 1-87, Spring 2008.
• The Easterlin “Paradox”: Within countries at a point in time, income and subjective well-being (SWB) are positively correlated: richer people are happier. This relationship between income and SWB does not seem to hold when looking between countries, or in data over long periods of time within a country.
• One possible resolution of the Easterlin Paradox is that reference-dependent preferences might be at work. People are happier if they are relatively rich within their country, but they don’t compare themselves with people in other countries or people thirty years ago. If happiness depends on relative income, maybe we need to have highly progressive taxation, as higher income (or work effort) for one person “imposes” a relative deprivation cost upon everyone else.
• Stevenson and Wolfers re-examine the evidence, and the Easterlin Paradox disappears: time series and cross-country studies display essentially the same correlation between SWB and (log) income as is calculated from within-country cross-sections. On average, higher income goes with higher SWB.
• Life satisfaction and happiness are not identical concepts, with happiness more about affect (in Kahneman's typology, System 1). Happiness is less strongly correlated than is SWB with income. Tanzania and Nigeria exhibit high happiness, low SWB.
• A rise in income of $100 contributes more to happiness in poor countries than in rich countries – but it contributes to higher happiness in both rich and poor countries.
• The US is the exception, with a small fall in SWB between 1972 and 2006, despite rising average income. The US data might reflect stagnation in middle-class income.
• The Easterlin “Paradox”: Within countries at a point in time, income and subjective well-being (SWB) are positively correlated: richer people are happier. This relationship between income and SWB does not seem to hold when looking between countries, or in data over long periods of time within a country.
• One possible resolution of the Easterlin Paradox is that reference-dependent preferences might be at work. People are happier if they are relatively rich within their country, but they don’t compare themselves with people in other countries or people thirty years ago. If happiness depends on relative income, maybe we need to have highly progressive taxation, as higher income (or work effort) for one person “imposes” a relative deprivation cost upon everyone else.
• Stevenson and Wolfers re-examine the evidence, and the Easterlin Paradox disappears: time series and cross-country studies display essentially the same correlation between SWB and (log) income as is calculated from within-country cross-sections. On average, higher income goes with higher SWB.
• Life satisfaction and happiness are not identical concepts, with happiness more about affect (in Kahneman's typology, System 1). Happiness is less strongly correlated than is SWB with income. Tanzania and Nigeria exhibit high happiness, low SWB.
• A rise in income of $100 contributes more to happiness in poor countries than in rich countries – but it contributes to higher happiness in both rich and poor countries.
• The US is the exception, with a small fall in SWB between 1972 and 2006, despite rising average income. The US data might reflect stagnation in middle-class income.
Sunday, October 25, 2015
Maniadis, Tufano, and List (2014) on Anchoring Effects
Zacharias Maniadis, Fabio Tufano, and John A. List, “One Swallow Doesn’t Make a Summer: New Evidence on Anchoring Effects.” American Economic Review 104(1): 277–290, 2014 [pdf of earlier version here].
• A 2003 article by Ariely, Loewenstein, and Prelec found huge anchoring effects in the willingness-to-pay for goods and the willingness-to-accept payment for listening to aversive sounds. These experiments, along with others, undermine the usual economics assumption of fixed preferences.
• Maniadis, Tufano, and List replicate parts of the earlier study and find some anchoring, but the effects are about ½ to ⅓ the size identified in the earlier article.
• The general conclusion on anchoring drawn by Maniadis, Tufano, and List: anchoring effects are real, but there is no evidence concerning their magnitude in economically important settings.
• Beware of interesting new findings! Statistical significance means little in isolation. Be a Bayesian: if the prior probability of a result is small, do not rush to accept it on the basis of one published study.
• The greater the number of independent researchers who are working on a specific finding, the lower the likelihood that a single publication identifying the finding is correct.
• Research and other biases also can enter the picture; for instance, journals might systematically be less interested in publishing non-findings than (statistically significant) findings.
• The replicability of a study is the key to inculcating rational belief. One or two independent replications tend to greatly add to the probability that a finding is correct. [It’s a movement!; see http://replicationnetwork.com/.]
• A potentially useful reference is Samuel Arbesman, The Half-Life of Facts: Why Everything We Know has an Expiration Date, Current, 2012.
• A 2003 article by Ariely, Loewenstein, and Prelec found huge anchoring effects in the willingness-to-pay for goods and the willingness-to-accept payment for listening to aversive sounds. These experiments, along with others, undermine the usual economics assumption of fixed preferences.
• Maniadis, Tufano, and List replicate parts of the earlier study and find some anchoring, but the effects are about ½ to ⅓ the size identified in the earlier article.
• The general conclusion on anchoring drawn by Maniadis, Tufano, and List: anchoring effects are real, but there is no evidence concerning their magnitude in economically important settings.
• Beware of interesting new findings! Statistical significance means little in isolation. Be a Bayesian: if the prior probability of a result is small, do not rush to accept it on the basis of one published study.
• The greater the number of independent researchers who are working on a specific finding, the lower the likelihood that a single publication identifying the finding is correct.
• Research and other biases also can enter the picture; for instance, journals might systematically be less interested in publishing non-findings than (statistically significant) findings.
• The replicability of a study is the key to inculcating rational belief. One or two independent replications tend to greatly add to the probability that a finding is correct. [It’s a movement!; see http://replicationnetwork.com/.]
• A potentially useful reference is Samuel Arbesman, The Half-Life of Facts: Why Everything We Know has an Expiration Date, Current, 2012.
Meub and Proeger (2014), Are Groups Less Behavioral?
Lukas Meub and Till Proeger, “Are Groups 'Less Behavioral'? The Case of Anchoring,” February 18, 2014; updated version available here.
• Do groups (as opposed to individuals) filter out irrationality and promote rationality? Do they “debias”? The broad consensus position is “yes,” though perhaps the advantage of a group is more pronounced with issues that have objectively correct answers (as opposed to being matters of judgment.)
• Experiments conducted by Meub and Proeger used groups of three players, faced with 15 questions, 5 each from the categories of: factual knowledge; likelihoods; and, price estimation. The experiments involved two conditions, the “calibration” condition and the “anchor” condition.
• The calibration condition is employed to identify appropriate anchors. The high anchor and low anchor values come from the 15th and 85th percentiles of unanchored responses. There is a small monetary incentive to guess well; payments varied (based on performance and condition) from 5 to 15 euro.
• The anchor index summarizes how much the anchors sway responses. The numerator of the index is the median difference in responses between the high anchor group and the low anchor group. The denominator is the difference between the high and low anchor values themselves. An anchor index of zero indicates that people gave the same response, on average, no matter what anchor they were exposed to. An anchor index of one indicates that the respondents, on average, just let the anchor be their response.
• The authors find that over all 15 questions, groups are less biased, with an anchor index of .34 as opposed to .52 for individuals.
• High anchors seem to matter more, but in any event, groups are less biased than individuals on the factual knowledge questions. Groups showed no advantage over individuals on the probability questions, nor for price estimation.
• Participants were asked if they perceived an anchoring problem, and the responses did not vary between individuals or groups.
• Do groups (as opposed to individuals) filter out irrationality and promote rationality? Do they “debias”? The broad consensus position is “yes,” though perhaps the advantage of a group is more pronounced with issues that have objectively correct answers (as opposed to being matters of judgment.)
• Experiments conducted by Meub and Proeger used groups of three players, faced with 15 questions, 5 each from the categories of: factual knowledge; likelihoods; and, price estimation. The experiments involved two conditions, the “calibration” condition and the “anchor” condition.
• The calibration condition is employed to identify appropriate anchors. The high anchor and low anchor values come from the 15th and 85th percentiles of unanchored responses. There is a small monetary incentive to guess well; payments varied (based on performance and condition) from 5 to 15 euro.
• The anchor index summarizes how much the anchors sway responses. The numerator of the index is the median difference in responses between the high anchor group and the low anchor group. The denominator is the difference between the high and low anchor values themselves. An anchor index of zero indicates that people gave the same response, on average, no matter what anchor they were exposed to. An anchor index of one indicates that the respondents, on average, just let the anchor be their response.
• The authors find that over all 15 questions, groups are less biased, with an anchor index of .34 as opposed to .52 for individuals.
• High anchors seem to matter more, but in any event, groups are less biased than individuals on the factual knowledge questions. Groups showed no advantage over individuals on the probability questions, nor for price estimation.
• Participants were asked if they perceived an anchoring problem, and the responses did not vary between individuals or groups.
Kahneman (2011), Chapter 11, “Anchors”
Daniel Kahneman, “Anchors.” Chapter 11, pages 119-128, in Thinking, Fast and Slow, New York: Farrar, Straus and Giroux, 2011.
• Anchoring occurs when a person considers one number, before estimating another, unknown number. The estimate will tend to stay close to the original number considered, even if that original number is known to be completely unrelated to the estimated quantity.
• Anchoring might result from two (separate?) mechanisms. One is priming, the power of suggestion, which works on our rapid responses (“System 1”); the second is when we deliberate (“System 2”) over the estimate by adjusting away from the anchor, but do not adjust sufficiently.
• The asking price for a house can be a form of anchor – one that affects professional real estate agents as well as amateurs. The amount of damages requested in a civil suit also might serve as an anchor. Further, legislated maxima (caps) on damage awards might serve as an anchor, drawing upwards what otherwise might have been smaller awards.
• Instructions to deliberately counter the anchor can be effective.
• Anchoring occurs when a person considers one number, before estimating another, unknown number. The estimate will tend to stay close to the original number considered, even if that original number is known to be completely unrelated to the estimated quantity.
• Anchoring might result from two (separate?) mechanisms. One is priming, the power of suggestion, which works on our rapid responses (“System 1”); the second is when we deliberate (“System 2”) over the estimate by adjusting away from the anchor, but do not adjust sufficiently.
• The asking price for a house can be a form of anchor – one that affects professional real estate agents as well as amateurs. The amount of damages requested in a civil suit also might serve as an anchor. Further, legislated maxima (caps) on damage awards might serve as an anchor, drawing upwards what otherwise might have been smaller awards.
• Instructions to deliberately counter the anchor can be effective.
Subscribe to:
Posts (Atom)