The IRS is quietly the most important regulator in the US. IRS audits can be a quiet, non-flashy way to drown political opponents (and firms) with costs. We don’t know very much about how the IRS operates given its importance imo. - ThreadSky

economeash.bsky.social • 101 days ago

The IRS is quietly the most important regulator in the US. IRS audits can be a quiet, non-flashy way to drown political opponents (and firms) with costs. We don’t know very much about how the IRS operates given its importance imo.

Reposted from Hetan Shah

Very good point here from @edwardluce.bsky.social - the modern and relatively invisible way to harass your opponents is through unending tax investigations that lead you to bankruptcy on.ft.com/4hGUJJR

Comments

menzer.bsky.social•101 days ago

What? We know a TON about how the IRS operates. A bunch of external researchers work with them and they have actually been really good about allowing research which does not always paint them favorably...

economeash.bsky.social•101 days ago

I think I disagree. This is pretty subjective (& I didn’t mean to sound reductive about research using IRS data) but the IRS is important as an administrator, regulator, and for law enforcement - I think there’s a lot more to do in this space but access remains a huge barrier (maybe for good reason)

economeash.bsky.social•101 days ago

Compare the IRS with the SEC or the PCAOB - I’d say IRS has significantly wider scope & impact but not noticeably more amount of research. At least in corporate tax, there aren’t many ways to get institutional access. Most “IRS scrutiny” papers use 10-K mentions & downloads - which only go so far

menzer.bsky.social•101 days ago

There is a lot of research on the SEC and PCAOB, but much of that is not done with internal agency data. It is external or requires FOIA requests. Which is quite different than the agency allowing researchers in.

menzer.bsky.social•101 days ago

As someone who works with the IRS and is currently going through the process to get access to work with internal PCAOB data. I would say the PCAOB is actually more restrictive at least in the initial phases.

economeash.bsky.social•101 days ago

Interesting - I would have guessed otherwise. Also sounds like you have more experience working with the IRS directly, my prior was that there were few teams that were let in & being in the good graces of the agency was essential for continued access. Happy to be corrected

menzer.bsky.social•101 days ago

I would also not say that "most" papers use IRS scrutiny. Very few do. ETRs, Distance to IRS office, A ton of different UTB measures, some cool stuff recently on textual analysis of 10-k MD&A

The 10-k downloads have actually gone out of favor because you can no longer de-anonymize the IP addresses

menzer.bsky.social•101 days ago

Access is limited by law (re: §6103), so any outside access for research purposes must have benefit tax administration in some way.

Is there more to do? There always is. But I would struggle to name another agency that lets outsiders study the agency's own processes and publish negative results

economeash.bsky.social•101 days ago

Also we’ve been talking about corporate tax for the most part, where the IRS is less likely to receive bad publicity regardless of the findings. My original post was about political connections, tax-exempt stuff etc., which I’ve seen very little if any research that uses IRS data.

menzer.bsky.social•101 days ago

I think my go-to example of that would be recent papers on racial bias in audits. This one got a lot of press, but there have been a few. (https://siepr.stanford.edu/publications/working-paper/measuring-and-mitigating-racial-disparities-tax-audits)

I have also seen some on things like customer service (albeit that was an experiment to try and make it better)

menzer.bsky.social•101 days ago

I think that some of the less flashy stuff that is just about improving the service does not make it to top journals. It is sometimes by internal economists/researchers, and so although it gets put out into the public, publication isn't the goal or point, and it gets buried in GAO or TIGTA reports

Comments

Posting Rules

Reply