site banner

Culture War Roundup for the week of January 30, 2023

This weekly roundup thread is intended for all culture war posts. 'Culture war' is vaguely defined, but it basically means controversial issues that fall along set tribal lines. Arguments over culture war issues generate a lot of heat and little light, and few deeply entrenched people ever change their minds. This thread is for voicing opinions and analyzing the state of the discussion while trying to optimize for light over heat.

Optimistically, we think that engaging with people you disagree with is worth your time, and so is being nice! Pessimistically, there are many dynamics that can lead discussions on Culture War topics to become unproductive. There's a human tendency to divide along tribal lines, praising your ingroup and vilifying your outgroup - and if you think you find it easy to criticize your ingroup, then it may be that your outgroup is not who you think it is. Extremists with opposing positions can feed off each other, highlighting each other's worst points to justify their own angry rhetoric, which becomes in turn a new example of bad behavior for the other side to highlight.

We would like to avoid these negative dynamics. Accordingly, we ask that you do not use this thread for waging the Culture War. Examples of waging the Culture War:

  • Shaming.

  • Attempting to 'build consensus' or enforce ideological conformity.

  • Making sweeping generalizations to vilify a group you dislike.

  • Recruiting for a cause.

  • Posting links that could be summarized as 'Boo outgroup!' Basically, if your content is 'Can you believe what Those People did this week?' then you should either refrain from posting, or do some very patient work to contextualize and/or steel-man the relevant viewpoint.

In general, you should argue to understand, not to win. This thread is not territory to be claimed by one group or another; indeed, the aim is to have many different viewpoints represented here. Thus, we also ask that you follow some guidelines:

  • Speak plainly. Avoid sarcasm and mockery. When disagreeing with someone, state your objections explicitly.

  • Be as precise and charitable as you can. Don't paraphrase unflatteringly.

  • Don't imply that someone said something they did not say, even if you think it follows from what they said.

  • Write like everyone is reading and you want them to be included in the discussion.

On an ad hoc basis, the mods will try to compile a list of the best posts/comments from the previous week, posted in Quality Contribution threads and archived at /r/TheThread. You may nominate a comment for this list by clicking on 'report' at the bottom of the post and typing 'Actually a quality contribution' as the report reason.

13
Jump in the discussion.

No email address required.

New Frontiers in Algorithmic Racism - Tax Edition

The New York Times has an article out on the IRS algorithmically targeting black Americans at higher rates than other racial groups. The claim is that there's something in the algorithm that inappropriately biases it against black Americans. Summarized in the opening paragraphs:

Black taxpayers are at least three times as likely to be audited by the Internal Revenue Service as other taxpayers, even after accounting for the differences in the types of returns each group is most likely to file, a team of economists has concluded in one of the most detailed studies yet on race and the nation’s tax system.

The findings do not suggest bias from individual tax enforcement agents, who do not know the race of the people they are auditing. They also do not suggest any valid reason for the I.R.S. to target Black Americans at such high rates; there is no evidence that group engages in more tax evasion than others.

OK, so what exactly is causing them to get audited more if it's not individual bias, the machines are blinded to the race of the individual, and the rules are the same for everyone? Apparently some of it comes down to targeting EITC filings:

Black Americans are disproportionately concentrated in low-wage jobs. They are more likely than whites to claim the E.I.T.C. The authors wondered if that prevalence in claiming the credit might explain why Black taxpayers face more audits, because I.R.S. data show the agency audits people who claim the E.I.T.C. at higher rates than other taxpayers.

But as the research progressed, the authors found the share of Black Americans claiming the E.I.T.C. only explained a small part of the audit differences. Instead, more than three-quarters of the disparity stems from how much more often Black taxpayers who claim the credit are audited, compared with E.I.T.C. claimants who are not Black.

Unless I'm missing something, the article does not explicitly state what the relevant factors are that result in this targeting are. In what I see as typical NYT style, it does leave a breadcrumb that might be suggestive if you're ignoring the narrative quotes embedded in the article:

Black taxpayers appear to disproportionately file returns with the sort of potential errors that are easy for I.R.S. systems to identify, like underreporting certain income or claiming tax credits that the taxpayer does not qualify for, the authors find.

To me, this reads like the most likely explanation for black taxpayers being audited more frequently is that they report their income incorrectly in easy-to-detect ways. Since the IRS already has W-2 data for filers, it's probably not very hard for them to notice when someone reports their income wrong. There isn't really any elaboration that I find after this, so I'm unclear on how much this accounts for auditing disparities. The implication of the article and the quotes from "equity" advocates imply to me that we should figure out a way to make sure that white Americans are audited at least as much as black Americans, regardless of who is misreporting their income more frequently.

As cynical as it sounds, I'm beginning to hear the term "algorithmic bias" as nothing more than a form of projection - algorithm systems frequently detect something real about the world, people with racially motivated politics don't like that outcome, and they seek to shift the algorithm towards a bias in favor of their preferred group. If a program that is optimized for detecting incorrect tax filings works as intended to detect them, but turns up more black Americans than white Americans, the suggestion appears to be to change the weighting until it evens out the races, regardless of the impact on the efficiency of detecting lost revenue. The "algorithmic bias", from my reading of this would be injecting a deliberate racial preference to counter the program noticing actual disparities. I am reminded of the racial resentment scale, in which people who say that "blacks have gotten less than they deserve" are not racially resentful, while those who think things like "Irish, Italian, and Jewish ethnicities overcame prejudice and worked their way up, Blacks should do the same without any special favors" are racially resentful.

Anyway, I'll be curious to see if the study is released more publicly and details what exactly is causing the disparity.

Expecting any sort of consistency is a fool's errand. The left vacillating on being pro-DHS/FBI during trump abut anti-DHS/FBI during Bush. Or pro-IRS during Obama. Both sides do it, so it's not just to pick on the left. The leopard does not care whose face it is that gets eaten. I think these organizations have too much power, and that bias is secondary to this.

As cynical as it sounds, I'm beginning to hear the term "algorithmic bias" as nothing more than a form of projection - algorithm systems frequently detect something real about the world, people with racially motivated politics don't like that outcome, and they seek to shift the algorithm towards a bias in favor of their preferred group. If a program that is optimized for detecting incorrect tax filings works as intended to detect them, but turns up more black Americans than white Americans, the suggestion appears to be to change the weighting until it evens out the races, regardless of the impact on the efficiency of detecting lost revenue. The "algorithmic bias", from my reading of this would be injecting a deliberate racial preference to counter the program noticing actual disparities. I am reminded of the racial resentment scale, in which people who say that "blacks have gotten less than they deserve" are not racially resentful, while those who think things like "Irish, Italian, and Jewish ethnicities overcame prejudice and worked their way up, Blacks should do the same without any special favors" are racially resentful.

A solution could be transparency, but if people knew how the the algorithms worked, like what triggers an audit, they would be gamed and rendered infective.

How would they game a fraud detection algorithm? By not committing fraud?

The classic example would be the old 10,000USD deposit at a bank triggering a reporting requirement, those reports focusing attention and investigation into one's finances and also slowing things on the customers end. Depositing 5,000USD and then later depositing 5,000USD does not trigger those reports and sometime between 1970 and 1986 there may have been common advice to do just that for convenience's sake. Of course, specifically depositing money in that way with the intent to avoid that sort of detection is now a federal felony. Often times many of the detection algorithms that have to be run by people end up as straight forward rules of if-this then-that so avoiding triggering detection in the common case might not be that difficult.

"Structuring" (breaking up a deposit into smaller deposits to avoid reporting) being a crime infuriates me. This is another aspect of the war on drugs seeping into financial regulation and corrupting the rules. In another horrifying example, the IRS is trying to find someone $2.1 million for failing to file a disclosure form. https://reason.com/2023/01/23/supreme-court-declines-case-challenging-excessive-irs-penalties/ No crimes were alleged, it wasn't drug money, the IRS just wants to know if you have a foreign bank account with more than $10k in it and if you don't file the form, they can take half the money in it. It's terrible.

I'm annoyed at the reporting requirements too, but the mirror image of money laundering is tax evasion, and governments are very motivated to prevent tax evasion by any means possible, up to and including totalitarian monitoring of all money flows.

With respect to the specific requirements to report foreign accounts: the reporting requirement is clearly stated in tax instructions and up to a few years ago the IRS was remarkably lax about requiring people (with less than $50,000 in their accounts) to report on time. The form for reporting foreign accounts even included checkboxes where one could state one's "reason for reporting late": "I forgot" and "I didn't know I had to" were valid options.

Granted, I'm still a bit confused by the reporting requirements and process for large wire transfers.

governments are very motivated to prevent tax evasion by any means possible, up to and including totalitarian monitoring of all money flows

Absolutely. That motivation is why any hope of a non-totalitarian end state requires strong pushback on this kind of thing. "Money laundering" is the "think of the children" of financial regulation. If one could report drug sales as "miscellaneous goods" there would be no reason to go through all of the hoops of washing the money. If all the government cared about was tax evasion, it would allow an amnesty category to report any income one didn't want to specify. Instead, the tax department has been roped into the criminal enforcement department and it makes for ridiculous regulations that shouldn't apply to 90% of the population.

You can report drug dealing as miscellaneous income. The problem is you're not allowed to deduct the costs of doing illegal business, so it's really not practical; you'd have to pay full income tax on the gross.