-
-
Notifications
You must be signed in to change notification settings - Fork 159
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
Update Tax-Calculator with latest CPS file and CPS weights #2444
Conversation
Also cc @MaxGhenis as a frequent user of the file. |
I'd like to wait on merging this until @andersonfrailey has an opportunity to take a look at the change in results. |
Missed when this was opened. I'll take a look later tonight |
@Peter-Metz I wouldn't worry about the changes in benefit and tax information primarily because we have validation scripts in taxdata to ensure that all income, benefit, and tax unit composition information is accurately represented in the final CPS file. Also, one of the steps in producing the old CPS file was splitting up any tax units that had information that was top-coded into 10 or 15 separate units and randomly changed each tax unit's value for the top coded variable, which would affect unweighted totals. We no longer do this because now instead of putting a hard cap on the variables that are subject to top coding, Census just shuffles them between similar people. As for any changes the weighted totals, we switched to a new solver for creating the weights, which would explain the differences there. |
@Peter-Metz, thanks for this PR and sorry that it now has a minor conflict. I added a WIP tag while we wait for that to be resolved. Afterwards, based on @andersonfrailey's feedback, it sounds like this is good to go. |
One more question before I resolve the merge conflicts -- several functions that calculate child tax credits in |
@Peter-Metz, this is a good point about I think it means we have two options:
I am inclined towards (1), but we could get away w (2) given that we are coming up on a major release. I think (1) is better since the Clinton reform is important as an historical artifact, and, if she proposed a reform depending on |
(I am seeing that I suggested (2) previously. PSLmodels/taxdata#329 (comment)) |
Kindergarten entrance is most typically based on whether you are 5 in September (3/4 of the way through the year), so I can see a stronger case for policies to use nu06 rather than nu05. https://nces.ed.gov/programs/statereform/tab5_3.asp |
I'm also seeing that @martinholmer suggested replacing Upon further consideration, I suggest we go ahead and do that as part of following (2) from #2444 (comment). So the difference from (2) is that we archive instead of amending the Clinton proposal. If someone wants |
@MattHJensen I agree with your reasoning to replace |
Codecov Report
@@ Coverage Diff @@
## master #2444 +/- ##
=======================================
Coverage 99.92% 99.92%
=======================================
Files 13 13
Lines 2582 2582
=======================================
Hits 2580 2580
Misses 2 2
Continue to review full report at Codecov.
|
The GH action to build the jb docs is currently failing while trying to execute Is it possible we need to activate the cc @jdebacker @MaxGhenis @hdoupe
|
Looks more likely to be a |
The latest commit changes cc @MattHJensen |
@Peter-Metz, could you remove the WIP tag when this is ready for another review? |
@MattHJensen done |
Thanks @Peter-Metz. Merging. |
This PR brings Tax-Calculator up to speed to the recent updates to the CPS file and CPS weights in PSLmodels/taxdata#314. In addition to changes to the files themselves,
nu05
was replaced withnu06
.A few notes:
cc @MattHJensen
Note: the
puf.csv
file (needed to pass tests locally) used for this PR was generated with the taxdata Makefile (i.e.make puf-files
). As of 8/6, the updatedpuf.csv
has not been distributed to users.