NEW UPDATE: Completed trials may now upload and register supplementary documents (e.g. null results reports, populated pre-analysis plans, or post-trial results reports) in the Post Trial section under Reports, Papers, & Other Materials.
Initial registration date
May 22, 2019
June 16, 2020 2:23 PM EDT
This section is unavailable to the public. Use the button below
to request access to this information.
Other Primary Investigator(s)
Additional Trial Information
In this project, I study how well people understand and how well they can learn about five economic policies: i) Personal income taxation, ii) Estate taxation, iii) Health insurance, iv) Trade, and v) Monetary policy. To that end, I run large-scale online surveys and experiments on representative U.S. samples and ask respondents a series of questions designed to get a picture of not only their factual knowledge about policies, but also their knowledge of the underlying economic phenomena, their understanding of the mechanisms of each policy, in particular its efficiency and distributional implications, and what first-order considerations come to their mind if they are prompted to think about the policy, without however nudging them to think about a particular effect or another. The latter is done with open-ended questions that are subjected to textual analysis, namely topic analysis and sentiment analysis. Respondents are also randomly asked to think about the effects of policies on themselves and how they would respond to the policies, as opposed to a general higher income or middle-class person. In addition, respondents are randomly asked to think about the effects of the policies on women more specifically.
Also experimentally, I show people instructional videos that explain the workings and consequences of each policy, but from different perspectives. The ``Distributional'' perspective focuses on the distributional consequences of each policy; the ``Efficiency'' perspective focuses on the efficiency costs; the ``Economist'' perspectives focuses on the trade-off, combining both the distributional and efficiency perspectives together. Registration Citation
We administer an online survey to respondents in the United States, eliciting their knowledge about the effects various economic policies. We also seek to understand respondents' understanding of the underlying economic mechanisms, their perceptions of fairness and their reactions to informational prompts and differentiated framings of questions.
Intervention Start Date
Intervention End Date
Primary Outcomes (end points)
The key outcome variables of the experiment are all the views on fairness, the knowledge of policies and understanding of economic mechanisms, as well as support for and views of different government policies.
Primary Outcomes (explanation)
We will use a lot of the policy knowledge and perception questions as such, but we also construct various indexes that summarize the support for government intervention, the perception of the efficiency costs of economic policies, the fairness of redistribution, the preference for open borders and so on. We will also construct measures of the misperceptions by subtracting respondents' answers to knowledge questions from the true answer, as well as variables indicating whether a respondent is particularly accurate or inaccurate. In the text analysis, we will construct measures of "topics" and "sentiment" using text analysis tools.
Secondary Outcomes (end points)
In addition to the experimental variables, we are also interested in the heterogeneity of the background characteristics of respondents and will thus look at outcomes by sub-groups as defined by income, age, gender, political affiliation, education level and field, main sources of news.
Secondary Outcomes (explanation)
We randomize 1) the information provided to respondents; 2) the framing of the questions; 3) the financial incentives provided for correct answers. For instance, some respondents are asked how they would react to policy changes, or how women in general would react, while others are asked how people in general would react.
Experimental Design Details
The randomization is done by the survey software (Qualtrics).
The unit of randomization is the individual respondent.
Was the treatment clustered?
Sample size: planned number of clusters
The planned number of clusters is around 15,000 individuals.
Sample size: planned number of observations
Approximately 15,000 individuals.
Sample size (or number of clusters) by treatment arms
Approximately half the sample sees no information video. Conditional on not seeing a video, respondents are evenly randomized between the framing branches ("you"/"gender"/neutral branches).
Minimum detectable effect size for main outcomes (accounting for sample design and clustering)
INSTITUTIONAL REVIEW BOARDS (IRBs)
IRB Approval Date
IRB Approval Number