Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add original weight to tax_microdata_<year>.csv.gz files #44

Closed
donboyd5 opened this issue Apr 22, 2024 · 0 comments · Fixed by #52
Closed

Add original weight to tax_microdata_<year>.csv.gz files #44

donboyd5 opened this issue Apr 22, 2024 · 0 comments · Fixed by #52

Comments

@donboyd5
Copy link
Collaborator

The tax_microdata_.csv.gz files currently have the final weight in the s006 column. This makes sense, but it would be good to add the original weight to the file as well - e.g., s006_original or something similar. This would have two benefits:

  1. Allows for direct analysis of how weights changed with reweighting - e.g., allowing calculation of percentiles of ratio of new weight to original weight, and similar analyses.

  2. Allows analysis of how "innocent bystanders" were affected by reweighting. For example, suppose we target number of returns and total AGI by AGI range but do not target potential SALT deductions or something else that may not be related to AGI in a simple way. That untargeted variable (an innocent bystander) may have been jerked around substantially in an effort to minimize the deviation from originally-weighted AGI by AGI range and the target AGI. And that change in SALT could, conceivably, move further away from what we want. Having the original weights will help us understand what damage we have done to untargeted variables when we adjust weights to move toward targets, and could suggest areas where we want to focus additional targeting. It will also help us understand the importance of anchoring revised weights so that they stay near original weights (that is, it will help us understand the importance of penalizing changes in weights).

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging a pull request may close this issue.

1 participant