Hacker News Reader

Comment by Twirrim | Hacker News Reader

One of the craziest things is, it save their tax department money for every single tax filing done online vs paper.

Paper takes an obscene amount of human hours to process in comparison. Online tax filings should at the very least be free, or even give you an extra tax deduction.

View on HN

> One of the craziest things is, it save their tax department money for every single tax filing done online vs paper

One of the HUGE benefits of e-file for them is that there are sanity checks that happen almost immediately and if they fail the IRS completely rejects the return. This does two things:

1. ensures accepted returns look much closer to accurate

2. places the burden of rejection notification on the e-file transmitter. For example, if I transmit a tax return that the IRS rejects due to a re-used SSN, the onus is now on me to notify that filer that their SSN has likely been stolen and provide customer service on what they should do next. From the IRS perspective this is a huge win. For e-file transmitters and tax software providers this is an un-fun problem to deal with.

> "IRS rejects due to a re-used SSN"

Would they really do that? It seems like that would give you two problems. Your SSN was stolen and now you can't file your taxes too.

The proper response would probably be to notify you, accept your documents (and the other documents presumably submitted by somebody else) tentatively, then do some investigation to determine which one is authentic.

Yes, indeed, they really do that. The person who has had their e-file reject must submit a paper return and call a special 800 number dedicated to this issue. It's a total double-whammy for the user. IRS tries to resolve these issues within 180 days... and if that delay of their return causes a financial burden they can file a petition. The wheels of progress turn slowly.

This is how a colleague of mine caught identity fraud. Someone had filed several fake tax returns with his information, called to have all of them invalidated as fake before submitting the "real" fake return, collecting the payment, and thus making it yet-harder for the real guy to claim the fake-but-not-fake was a false return.

Hmm. Canadian here, our SSN is called an SIN. NETFILE, which I didn't think was that innovative, matches SIN number to date of birth on file (which can be changed by phone call) and regardless, sends a refund to the address on file (which, like Direct Deposit settings, can be changed through a web interface). That said, I've never heard of someone committing tax fraud electronically, given the barrier to change an address. Google revealed no reported incidents, not even on forums. And if I had to submit on paper, as only 25% or so of Canadians do now, the software will let me print out all the forms I need, and if necessary, the PDFs from the CRA have fillable fields, IIRC. That said, there are still too many form types, even though most don't apply to me.

A lot of general phone support has been shut down and you are now told to look at the website.

I think the temporary phone shut down during the last debt ceiling government shutdown saga is now permanent.

OTOH, with this system, once you file you are protected from future identity theft. Since most people will file by the deadline, that limits the potential exposure.

In much of Northern Europe we use our digital IDs to file our taxes. Which means it is hard to submit fake tax returns in my name.

Not that crazy. This is seen and encouraged in startups all the time: Companies should charge for the value they provide, not for the man-hours they're spending.

Additionally, anything that makes things significantly easier probably requires a large up-front investment of man-hours and/or money. If you've ever dealt with government software projects, you can imagine how large of a project this must have been. That money has to be recouped somehow, and that won't happen by giving things away for free.

(Note: I'm not advocating for the sleazy behavior of lobbying against easier/cheap tax filing. I'm just making the case that there's business reasoning behind this, other than plain spite for the common folk.)

Do you know of actual/estimated numbers for human/computer processed tax filings? It would be interesting to see how many minutes of human time are needed for the average paper return. I have a sneaking suspicion that the "human touch time" is not "obscenely high."

It's subjective. But also I assume error rates are higher for those who file paper forms vs electronically.

I assume that error percentage is not the only difference between paper/e-filed returns.

IRS estimates an e-filed tax return will have the refund issued within 21 days. The paper form equivalent takes 6-8 weeks.

Source: http://www.irs.gov/taxtopics/tc152.html

The other thing to know is that e-filed tax returns are subject to an immediate sanity check (i.e., has this ssn already been used?, is this a valid zip code?), which doesn't exist for paper forms.

In Canada, 8 business days vs. 4-6 weeks: http://www.netfile.gc.ca/fq_rfnd-eng.html

For me, personally, it was 7 business days this year, and I did the filing in all of an hour and a half, in an afternoon. And that includes the time spent trying out two other free apps to decide which ones I liked better.

I appreciate the reply but I think time-to-refund is a terrible metric for estimating how much human touch time is involved in processing a return. What more do we know about human paper processing time per return than before?

How much would your answer change if you found out that there are only two people processing paper returns? 2^32 people? What if you found out that the types of returns that are mailed in are 10x more complicated/detailed than e-filed returns? Would it change if you knew there was an intentional processing delay on paper returns to increase e-file adoption?

I think it's a decent metric, and I'm surprised you don't agree. If we forget the "human time" part, it's a clear indicator of which method is more efficient.

Given that you want to "forget the 'human time' part" I am curious what you think the metric is for. Maybe you meant to respond to a different comment. In this thread I had asked if anyone had any numbers/estimates for processing time. I wanted these numbers because "It would be interesting to see how many minutes of human time are needed for the average paper return."

I had assumed "minutes of human time" was a proxy-metric for efficiency.

Reply on news.ycombinator.com