| Field | Before | After |
|---|---|---|
| Field Trial Title | Before Beliefs in Educational Meritocracy, Tolerance of Inequality, and Social Cohesion | After Sources of educational inequality and redistributive behavior: Experimental evidence |
| Field Abstract | Before We provide causal evidence on how beliefs in educational meritocracy affect tolerance of education-based inequality and attitudes towards the less-educated. Educational meritocracy is the idea that academic success indicates individual merit (e.g. hard work, effort, talent) rather than external circumstances (e.g. parental background, race, sex). We first elicit participants’ beliefs about how college attendance varies by socioeconomic status. Half of the participants will then receive the actual statistics, taken from Chetty et al. (2020). We then measure participants’ support for reducing income inequality between college and non-college graduates, attitudes towards the less-educated, support for educational reform policies, and willingness to donate to charities that promote access to higher education. | After We provide causal evidence on how beliefs about the sources of educational inequalities support for policies or initiatives that reduce education-related inequalities. We first elicit participants’ beliefs about how college attendance varies by socioeconomic status. Half of the participants will then receive the actual statistics, taken from Chetty et al. (2020). We then measure participants’ support for reducing income inequality between college and non-college graduates, support for educational reform policies, and willingness to donate to charities that promote access to higher education. This document outlines our plan for data analysis. |
| Field JEL Code(s) | Before C91, D63, D83, D91, I24 | After D63, D64, D83, I24 |
| Field Last Published | Before December 11, 2022 09:22 AM | After May 22, 2023 07:46 AM |
| Field Primary Outcomes (End Points) | Before (1) Preferred earnings gap between college and non-college graduates (2) Charitable donations (3) Support for policies that reduce barriers to attending college (4) Implicit attitudes towards less-educated | After (1) Charitable donations (2) Preferred earnings ratio between college and non-college graduates (ex-post redistribution) (3) Support for policies that reduce barriers to attending college (ex-ante policy support) |
| Field Primary Outcomes (Explanation) | Before (1) Preferred earnings gap between college and non-college graduates: Participants will be given the current income disparity between the average college graduate and average non-college graduate and asked what they think the income disparity between these two individuals should be. (2) Charitable donations: Participants will be told they have been automatically enrolled in a lottery for $100 and, if they win, they can choose to donate some (or all or none) of their winnings to a charity whose primary mission is to tackle inequalities in educational attainment at the tertiary level. (3) Support for policies that reduce financial barriers to attending college: Participants are provided with information on a policy that aims to reduce financial barriers to attending college (expanding the size of the Pell Grant, encouraging colleges to offer automatic application fee waivers for low-income students). They are then asked how much they would support the given policy. (4) Implicit attitudes towards less-educated: Participants take a novel version of the Implicit Association Test (IAT), designed for this study. This version of the IAT, designed for this study, assesses the ease with which participants make pleasant or unpleasant associations between typically white male names, which are either listed with or without an educational qualification (e.g. BSc, J.D., PhD). | After (1) Charitable donations: Participants will be told they have been automatically enrolled in a lottery for $100 and, if they win, they can choose to donate some (or all or none) of their winnings to a charity whose primary mission is to tackle inequalities in educational attainment at the tertiary level. (2) Preferred earnings ratio between college and non-college graduates (ex-post redistribution): Participants will be given the current income ratio between the average college graduate and average non-college graduate and asked what they think the income disparity between these two individuals should be. (3) Support for policies that reduce financial barriers to attending college (ex-ante policy support): Participants are provided with information on a policy that aims to reduce financial barriers to attending college (expanding the size of the Pell Grant, encouraging colleges to offer automatic application fee waivers for low-income students). They are then asked how much they would support the given policy. |
| Field Experimental Design (Public) | Before We first elicit participants’ beliefs about inequality in college attendance, measured by the percentage of 4-year-college attendees in a given birth cohort who grew up in each quintile of the income distribution. Using a between-subjects design, we then randomly allocate participants to a treatment or control group. The treatment group receives information about the true percentages of college attendees who grew up in households whose income is in the bottom or top quintile of the distribution. The control group receives information about college attendees that is unrelated to meritocracy in educational attainment. Both sets of information are computed using deidentified administrative data by Chetty et al. (2020). We then measure three self-reported outcomes (the preferred wage ratio between college and non-college graduates, attitudes towards the less-educated, support for policies that reduce financial barriers to attending college), and one real outcome (the choice to make an actual donation to an education-related charity). We also ask questions to understand the mechanisms through which the information treatment affects these outcomes. | After We first elicit participants’ beliefs about inequality in college attendance, measured by the percentage of 4-year-college attendees in a given birth cohort who grew up in each quintile of the income distribution. Using a between-subjects design, we then randomly allocate participants to a treatment or control group. The treatment group receives information about the true percentages of college attendees who grew up in households whose income is in the bottom or top quintile of the distribution. The control group receives information about college attendees that is unrelated to inequalities in educational attainment. Both sets of information are computed using deidentified administrative data by Chetty et al. (2020). We then measure two self-reported outcomes (the preferred wage ratio between college and non-college graduates and support for policies that reduce financial barriers to attending college), and one real outcome (the choice to make an actual donation to an education-related charity). We also ask questions to understand the mechanisms through which the information treatment affects these outcomes. |
| Field Planned Number of Clusters | Before 4200 individuals | After 2000 individuals |
| Field Planned Number of Observations | Before 4200 individuals | After 2000 individuals |
| Field Sample size (or number of clusters) by treatment arms | Before 2100 people in each treatment | After 1000 people in each treatment |
| Field Power calculation: Minimum Detectable Effect Size for Main Outcomes | Before 3780 participants will give us 0.8 power to detect an effect size of 0.10 of a standard deviation between the treatment and the control group in the main study at a .05 significance level. Therefore, the 4200 participants from the main study will give us more than 0.8 power to detect the same effect size. | After 1400 participants will give us 0.8 power to detect an effect size of 0.10 of a standard deviation between the treatment and the control group in the main study at a .05 significance level. Therefore, the 2000 participants from the main study will give us more than 0.8 power to detect the same effect size. |
| Field Additional Keyword(s) | Before Inequality, Meritocracy, Policy preferences, Survey experiment | After Attitudes toward inequality, Education, Policy preferences, Survey experiment |
| Field Public analysis plan | Before No | After Yes |
| Field | Before | After |
|---|---|---|
| Field Document | Before |
After
Educationism pre-analysis plan Jul2022.pdf
MD5:
e7f0de1c4a2fd765dea812411400aa5d
SHA1:
d188407f40dea41a3a6745b030424e6bae85c207
|