Ann-Kathrin Koessler, Benno Torgler, Lars P. Feld, and Bruno S. Frey,
“Commitment to Pay Taxes: A Field Experiment on the Importance of
Promise.” Tax and Transfer Policy Institute, Australian National University,
Working Paper 10/2016, November 29, 2016 [pdf here].
• A natural field experiment (n≈2000) is conducted in Switzerland (in 2013); the subjects, Swiss taxpayers, do not know that they are taking part in an experiment.
• The experiment concerns whether it is possible to encourage timely tax payments by having taxpayers voluntarily promise to remit their taxes on time. One potentially confounding factor, however, is that a new “dunning” policy for late taxpayers is enacted concurrently with the experiment.
• There are two “promise” treatments. In both cases, subjects are told that if they fill in and return a postcard promising to pay their taxes on time, and then do pay their taxes on time, they will be entered into a lottery. The differences between these two treatments is that in one case the lottery prize is cash – 1000 Swiss francs – and in the other, the prize is a wellness spa trip for two, worth approximately 1000 Swiss francs.
• Both promise treatments have parallel treatments that provide the same lottery to punctual payers, but that do not require or provide the option for the non-binding promise. A control treatment with no lottery or promise completes the collection.
• Almost one-third of the subjects who are given an opportunity to promise to pay on time make the promise. The willingness to promise is a pretty strong signal both of whether you have paid on time in the past, and of whether you will pay on time this year.
• Those who were in the spa lottery and who made the promise saw a significant jump in their compliance rates. But the lotteries with promise opportunities do not seem to do any better overall than the lotteries without the promise opportunities.
• My takeaway, perhaps not as optimistic as that of the authors, is that the “promise” intervention is pretty weak tea.
Sunday, January 28, 2018
Friday, January 5, 2018
Horton and Zeckhauser (2016) on Peer Effects in Production
John J. Horton and Richard J. Zeckhauser, “The Causes of Peer Effects in Production: Evidence from a Series of Field Experiments.” NBER Working Paper No. 22386, July 2016 [ungated pdf version here].
• A worker’s productivity is influenced by the productivity of her co-workers. Why?
• Do low-productivity workers fear punishment, because their slackerdom imposes more work on others, or because others simply view the provision of low effort as unfair? Do workers have a preference to not be unproductive? Does the performance of other employees signal to a worker the employer’s expectations?
• The field experiments involve hiring workers online (MTurk) to label images; the workers do not know that they are taking part in an experiment.
• Among the findings are that peers will punish slackers even when the slackers do not impose any harm on the other workers; and, that equity concerns motivate such punishment, through a suspicion of low effort from slackers. Workers therefore want either to avoid being slackers, or, perhaps, to avoid being perceived as slackers.
• Workers (the subjects in the experiments) label images and evaluate the labelling performance of other workers. Evaluations involve a recommendation as to whether the peer worker should be paid, and also, a suggestion of how a 9-cent bonus should be split between the evaluator and the evaluated. The recommended bonus splits are implemented. A recommendation that a worker not be paid, though not implemented, is taken to be a type of punishment of that worker.
• All of the subjects also evaluate the work of either a (specific) high effort or a low effort worker, based on work from a previous experiment. A description of the four treatments in the field experiment follows:
Baseline: Workers are shown either a very good labelling job, or a minimal effort version. They then choose to take the task (if indeed they want to), and start labelling.
Punish: After seeing a sample of excellent work, this treatment proceeds like Baseline, except that then some good or bad work requires peer evaluation.
Peers: After evaluation, subjects then are given a second labelling task. Will evaluating a good or a bad job (done by someone else) alter the worker’s second-round performance?
Explicit: This is like Peers, except that workers are told to produce only two labels per image. But when they evaluate others, some of the subjects see an image with the requisite two labels, while the others see an image with 11 labels (produced by some Stakhanovite). Note that this excessive production explicitly contravenes the injunction to provide only two labels.
• And the results...
Baseline: If you are shown a high-quality sample, you are less likely to take the job, but if you take it, you increase your effort.
Punish: Peer evaluation leads to lessened punishment for good as opposed to bad work; workers who themselves produce good work punish more.
Peers: If you evaluate a strong worker, your subsequent work is stronger, and the effect is more pronounced if you yourself are a high productivity worker.
Explicit: Workers shown against-the-rules overproduction do not punish it, and many workers themselves switch to excessive output.
• A worker’s productivity is influenced by the productivity of her co-workers. Why?
• Do low-productivity workers fear punishment, because their slackerdom imposes more work on others, or because others simply view the provision of low effort as unfair? Do workers have a preference to not be unproductive? Does the performance of other employees signal to a worker the employer’s expectations?
• The field experiments involve hiring workers online (MTurk) to label images; the workers do not know that they are taking part in an experiment.
• Among the findings are that peers will punish slackers even when the slackers do not impose any harm on the other workers; and, that equity concerns motivate such punishment, through a suspicion of low effort from slackers. Workers therefore want either to avoid being slackers, or, perhaps, to avoid being perceived as slackers.
• Workers (the subjects in the experiments) label images and evaluate the labelling performance of other workers. Evaluations involve a recommendation as to whether the peer worker should be paid, and also, a suggestion of how a 9-cent bonus should be split between the evaluator and the evaluated. The recommended bonus splits are implemented. A recommendation that a worker not be paid, though not implemented, is taken to be a type of punishment of that worker.
• All of the subjects also evaluate the work of either a (specific) high effort or a low effort worker, based on work from a previous experiment. A description of the four treatments in the field experiment follows:
Baseline: Workers are shown either a very good labelling job, or a minimal effort version. They then choose to take the task (if indeed they want to), and start labelling.
Punish: After seeing a sample of excellent work, this treatment proceeds like Baseline, except that then some good or bad work requires peer evaluation.
Peers: After evaluation, subjects then are given a second labelling task. Will evaluating a good or a bad job (done by someone else) alter the worker’s second-round performance?
Explicit: This is like Peers, except that workers are told to produce only two labels per image. But when they evaluate others, some of the subjects see an image with the requisite two labels, while the others see an image with 11 labels (produced by some Stakhanovite). Note that this excessive production explicitly contravenes the injunction to provide only two labels.
• And the results...
Baseline: If you are shown a high-quality sample, you are less likely to take the job, but if you take it, you increase your effort.
Punish: Peer evaluation leads to lessened punishment for good as opposed to bad work; workers who themselves produce good work punish more.
Peers: If you evaluate a strong worker, your subsequent work is stronger, and the effect is more pronounced if you yourself are a high productivity worker.
Explicit: Workers shown against-the-rules overproduction do not punish it, and many workers themselves switch to excessive output.
Tuesday, January 2, 2018
Hermann and Musshoff (2016) on Measuring Time Preferences
Daniel Hermann and Oliver Musshoff, “Measuring Time Preferences:
Comparing Methods and Evaluating the Magnitude Effect.” Journal of
Behavioral and Experimental Economics 65: 16-26, December 2016.
• Two different approaches to measuring discount rates in the past have revealed similar rates for US students – but what about for entrepreneurs, and for German students?
• This article consist of a web-based experiment with German farmers (n=111) and students (n=178); farmers are standing in for entrepreneurs, as farmers must make significant investment decisions that only yield results in the long-term.
• The experiments are conducted with both 100 and 300 euro benchmark amounts; the idea is to test for the “magnitude effect,” in which revealed discount rates fall as monetary amounts rise.
• Further, some previous estimates of discount rates might be skewed by the assumption of risk neutrality.
• The Coller and Williams (CW) task: You can receive €100 in three weeks. Or, you can receive more than €100 in twelve weeks. How much more than €100 do you need before you are willing to wait the extra nine weeks? This experiment is repeated for amounts 3 times as high.
• The Holt and Laury (HL) task: Lottery A offers prizes of either €180 or €144, while lottery B offers the outcomes €346.50 or €9. The higher prize has the same probability of occurring in both lotteries. How high does the probability of the higher prize have to be to get you to choose Lottery B?
• The p task, from Laury et al. (2012): Lottery A pays out zero half the time and €100 half the time, with the prize collected in three weeks. Lottery B pays out zero or €100 too, but doesn’t pay out until 12 weeks from now. How much greater than .5 does the probability of winning €100 have to be to get you to choose Lottery B, and thus wait the additional nine weeks? As with CW, this experiment is replicated with payouts three times as high.
• For farmers, the estimated average discount rate from CW is 12.9% for €100 tests, and 8.8% for €300 tests. For the p task, rates are significantly higher, at 30.6% for the €100 test, and 28.6% for the €300 test.
• In the joint estimation, student discount rates are similar to the farmers’. For the p-test on students, while this method still led to higher discount rates (significantly so for the €300 version) compared to the joint estimation, the increase was not nearly as a great as it was for farmers.
• For both students and farmers, raising the stakes to €300 lowers discount rates significantly in the joint estimation – the decline is greater for students. In the p-test approach, the fall in discount rates associated with higher stakes is not significant. This non-result suggests that for risky alternatives, the magnitude effect might not exist.
• Two different approaches to measuring discount rates in the past have revealed similar rates for US students – but what about for entrepreneurs, and for German students?
• This article consist of a web-based experiment with German farmers (n=111) and students (n=178); farmers are standing in for entrepreneurs, as farmers must make significant investment decisions that only yield results in the long-term.
• The experiments are conducted with both 100 and 300 euro benchmark amounts; the idea is to test for the “magnitude effect,” in which revealed discount rates fall as monetary amounts rise.
• Further, some previous estimates of discount rates might be skewed by the assumption of risk neutrality.
• The Coller and Williams (CW) task: You can receive €100 in three weeks. Or, you can receive more than €100 in twelve weeks. How much more than €100 do you need before you are willing to wait the extra nine weeks? This experiment is repeated for amounts 3 times as high.
• The Holt and Laury (HL) task: Lottery A offers prizes of either €180 or €144, while lottery B offers the outcomes €346.50 or €9. The higher prize has the same probability of occurring in both lotteries. How high does the probability of the higher prize have to be to get you to choose Lottery B?
• The p task, from Laury et al. (2012): Lottery A pays out zero half the time and €100 half the time, with the prize collected in three weeks. Lottery B pays out zero or €100 too, but doesn’t pay out until 12 weeks from now. How much greater than .5 does the probability of winning €100 have to be to get you to choose Lottery B, and thus wait the additional nine weeks? As with CW, this experiment is replicated with payouts three times as high.
• For farmers, the estimated average discount rate from CW is 12.9% for €100 tests, and 8.8% for €300 tests. For the p task, rates are significantly higher, at 30.6% for the €100 test, and 28.6% for the €300 test.
• In the joint estimation, student discount rates are similar to the farmers’. For the p-test on students, while this method still led to higher discount rates (significantly so for the €300 version) compared to the joint estimation, the increase was not nearly as a great as it was for farmers.
• For both students and farmers, raising the stakes to €300 lowers discount rates significantly in the joint estimation – the decline is greater for students. In the p-test approach, the fall in discount rates associated with higher stakes is not significant. This non-result suggests that for risky alternatives, the magnitude effect might not exist.
Monday, January 1, 2018
Wang, Rieger, and Hens (2017) on Culture and Loss Aversion
Mei Wang, Marc Oliver Rieger, and Thorsten Hens, “The Impact of Culture
on Loss Aversion.” Journal of Behavioral Decision Making 30: 270-281, 2017.
• Wang, Rieger, and Hens look at loss aversion in 53 countries. The underlying notion is that emotions are implicated in loss aversion, but the display and regulation of emotions is culturally influenced.
• In their data, higher (nationwide) levels of loss aversion are connected to: individualism; “power distance” (which means strength of social hierarchy, according to Wikipedia); and masculinity. A fourth factor, uncertainty avoidance, is less meaningful. These factors are drawn from a 2001 book on the consequences of culture by Geert Hofstede. [But it seems the subsequently expanded Hofstede criteria might be even more connected to loss aversion. Again, from Wikipedia: “Independent research in Hong Kong led Hofstede to add a fifth dimension, long-term orientation, to cover aspects of values not discussed in the original paradigm. In 2010, Hofstede added a sixth dimension, indulgence versus self-restraint.”]
• Loss aversion is measured by the answers to two questions. If you have a 50% chance of losing $25, and a 50% chance of winning x, how high does x have to be for you to agree to take this bet? The second question replaces $25 with $100. The loss aversion parameter is determined by dividing the answer by the stakes (25 or 100), and the overall measure averages the two stake-differentiated results.
• The survey is given to college students and the stakes are expressed in ways that, for students, are comparable across countries. Georgians have (easily) the highest loss aversion, at 7.5, with a few countries (Luxembourg, Bosnia, Tanzania) around 1 (no loss aversion). The mean across countries is 2.0. Eastern Europeans have the highest loss aversion, and Africans have the lowest.
• Women tend to be more loss averse, even though an increase in “masculinity” brings higher loss aversion. A greater percentage of Orthodox Christians leads to more loss aversion in a nation.
• Economic factors don’t seem to matter: it is culture, not the economy, that drives international differences in loss aversion.
• Wang, Rieger, and Hens look at loss aversion in 53 countries. The underlying notion is that emotions are implicated in loss aversion, but the display and regulation of emotions is culturally influenced.
• In their data, higher (nationwide) levels of loss aversion are connected to: individualism; “power distance” (which means strength of social hierarchy, according to Wikipedia); and masculinity. A fourth factor, uncertainty avoidance, is less meaningful. These factors are drawn from a 2001 book on the consequences of culture by Geert Hofstede. [But it seems the subsequently expanded Hofstede criteria might be even more connected to loss aversion. Again, from Wikipedia: “Independent research in Hong Kong led Hofstede to add a fifth dimension, long-term orientation, to cover aspects of values not discussed in the original paradigm. In 2010, Hofstede added a sixth dimension, indulgence versus self-restraint.”]
• Loss aversion is measured by the answers to two questions. If you have a 50% chance of losing $25, and a 50% chance of winning x, how high does x have to be for you to agree to take this bet? The second question replaces $25 with $100. The loss aversion parameter is determined by dividing the answer by the stakes (25 or 100), and the overall measure averages the two stake-differentiated results.
• The survey is given to college students and the stakes are expressed in ways that, for students, are comparable across countries. Georgians have (easily) the highest loss aversion, at 7.5, with a few countries (Luxembourg, Bosnia, Tanzania) around 1 (no loss aversion). The mean across countries is 2.0. Eastern Europeans have the highest loss aversion, and Africans have the lowest.
• Women tend to be more loss averse, even though an increase in “masculinity” brings higher loss aversion. A greater percentage of Orthodox Christians leads to more loss aversion in a nation.
• Economic factors don’t seem to matter: it is culture, not the economy, that drives international differences in loss aversion.
Subscribe to:
Posts (Atom)