Saez and Zucman tax plan calculator

Saez and Zucman have a tax plan creator on the website for their book, “The Triumph of Injustice.”

It lets users play with wealth, personal income, corporate, and estate taxes, and also change assumptions around healthcare.

It also includes presets for the Biden, Sanders, and Warren plans, as well as their own plan.

The output is a graph with the effective tax rate by percentile (and the top 400, like the NYT piece). I don’t believe it’s open-source like their is.


For those interested, the code for their tables and an Excel version of the tax plan creator can be downloaded from the technical appendix.

I’m currently exploring whether the data files on which their calculator is based are compatible with Tax-Calculator. I think they are, but don’t have it running yet. (@MaxGhenis, if you ahead of me, please let me know!)

I haven’t yet tried to plug their data into taxcalc, this would be great though.

This conversation between Saez, Summers, and Mankiw was fun.

Among much else, Summers discussed data quality and assumptions that are relevant to modeling a wealth tax.

Mankiw discussed the merits of a $1,000/mo UBI financed by a 20% tax on all income as compared to a $1,000/mo transfer with a 20% clawback financed by a 20% tax on income over $60,000.

Questions from Boushey, Acemoglu, and Furman.

Nice, Mankiw’s 2016 blog post on UBI vs. NIT remains my concise go-to.

@matthjensen Did SZ release their microdata? I’m looking through the files and only seeing summaries.

@MaxGhenis, is this what you are looking for?
Distributional national account micro-files (dta)

You are likely already aware, but I was pleasantly surprised to learn that Pandas has built in functions for reading .dta files: (HT to Lars Vilhuber for mentioning that .dta is an open format to me at some point.)

Also, PZ claim that the files for TaxJusticeNow are available in that online data appendix, but when I download the “programs and data for replication” the folder claiming to contain the data is empty.

Somewhere they say that you can ask for PUF-based dataset if you have access to the PUF. The DINA files I sent above as well as the files that are supposed to be in the TJN data appendix are grouped into bunches of 5 tax units. I’ve been planning to see if we can get the publicly available data working with Tax-Calculator before asking for PUF-based files.

The data SZ use for ( combines DINA with SCF and the Forbes 400. I believe they use this combined dataset for and their book’s analyses. It seems like they haven’t published this consolidated file.