Because the 3rd party company already exists and will provide integration (ie: ADP integration) that the IRS will not be able to provide... not without a few years of development or some sort of coercion.

At $50 for TurboTax (which is still more than the $20 I pay for TaxAct), its still an adequate solution that beats the current status quo.

--------

So my question to you is: do you not realize how bad the status quo actually is? The average preparation cost is $200 according to the IRS, which is very far above "TurboTax Deluxe edition", and far far above any of the other 3rd party tax solutions.

If people would prepare their own taxes with TurboTax, life would in fact be better than the current state. Period.

---------

Now we can imagine the IRS creating a website and everything will be Rainbows and Unicorns, but lets not forget about the Obamacare website. Turns out that competent web-developers are expensive, that government employees aren't competent web-developers, and finally aren't even that competent at figuring out contractors.

I can imagine a future where the US Government raises wages for once, builds out some competent project managers, retains talent, attracts talent, finds a contractor who can help them develop this hypothetical website, and then hope that the next $500-million+ web-project is actually successful.

Or... you know. I can just recommend people to use TurboTax (or TaxAct). Which already works, and is better than the status quo.

> Because the 3rd party company already exists and will provide integration (ie: ADP integration) that the IRS will not be able to provide... not without a few years of development or some sort of coercion.

You do realize that the IRS already gets a copy of your W2, 1099s, 1098s, etc, right? You can order up a transcript of the data that the IRS has on your file for a given year and they have a pretty complete set of data from all the various mandatory forms.

What they don't have (and can't have, because only you have them) is data you'd need for schedules C and E and much of the deductions for schedule A. Guess what? Intuit and your local "tax guy" don't have those either, unless you give it to them.

The IRS is not set up to provide people with their data during tax season. ADP can provide your year-end tax information before your W-2 is sent to you or the IRS. So, while I agree the IRS has the potential to use the data they already receive, the transformation of how they receive and use it is large enough that they can't be considered to have an existing solution in place.

Agreed they don't have a solution in place that can do the calculations by mid-April. They do have a solution in place that could do it by late summer, so if people just opted-in (even online), the IRS could take it from there, and send a refund or a bill by Labor Day. For people with simple tax situations (not me, alas), that would be a boon to people.

I worry more about the tax law changes that might come if people started paying EVEN LESS attention than today about how their taxes are calculated than I do about the IRS ability to execute this.

Clicking a webform and getting a bill or direct deposit divorces people from the fact that this is real money in play.

I'll give you that point.

But at the same time, the IRS has no real competitive alternative to something like TurboTax. Sponsoring the creation of yet another tax-preparation software feels like XKCD Comic number 927

http://xkcd.com/927/

who would you trust to write you a spiffy new tax app, CGI Federal or Facebook/Google? Has CGI every written anything remotely functional. They exist to be good at government paperwork and compliance, and then contract out the work. The process doesnt scream good software.

Given how damn good turbotax is, do you really trust the government to write something better?

Now if the government pledged to standardize API access between state and federal, and made it easy for startup companies to light a fire under incumbents, you get competition and government action/improvement.

> Given how damn good turbotax is, do you really trust the government to write something better?

Yes.

https://www.whitehouse.gov/digital/united-states-digital-ser...

https://18f.gsa.gov/

EDIT: I both know people at the USDS, and have interviewed with them. They are passionate about getting tech in government fixed. I believe (if Congress doesn't defund them), they'll deliver.

I know people in the government too. They're all passionate and loyal workers, often taking a job because they believe in the greater good as opposed to making money.

But 18f is a relatively small project (~100 employees), and overall USDS is relatively new. Veteran's Affairs for example STILL isn't digitized and is almost entirely a paper-run system. USDS is likely going to focus on that as opposed to building out a new website for the IRS.

Just realistically speaking, the kind of project you are proposing will take years to deliver version 1.0, and there's no guarantee that the USDS team will even handle it as there are definitely higher-priority tasks.

You are grossly overestimating the US Government's capabilities. I'm not lying when I say that the US Government needs to

1. Create a budget (we're still under continuing resolutions here. A budget hasn't been passed since like 2012)

2. Raise wages for once to attract talent (it has almost been a decade without wage growth, very bad for morale). FYI: the government shutdown threats (ie: federal employees / contractors don't get paid randomly) aren't helping.

3. Build up the IT teams, and I mean more than just a single 100-man team at 18F. The entirety of 18F, as good as they are, are way too small to handle a project on the scale of "Recreate Turbotax for All Americans... for free".

4. Build up project managers in the meantime who can handle contractors.

5. Hire competent contractors to build things out.

6. Execute all the above while Congress continues to refuse to pass a budget and the Republican half of America seems like they're going to lynch "greedy" Federal Workers who don't actually have great pension plans anymore (only the older Federal Workers have the crazy-good pension plan. Any millenial / college age kid who gets hired by the Feds don't really see as many benefits)

There are some fundamental problems here that need to be fixed before the US Government can tackle a project like you say successfully.

EDIT: one of the leading Presidential Candidates (Ted Cruz) is on the record for wanting to "Abolish" the IRS. People in general don't like Federal Workers, and that harms recruitment efforts. Which harms the kinds of project you want to start up. Its partially a self-fulfilling prophecy at this point.

I can't share details, because it was part of my interview process to be on the USDS team within the IRS, but the IRS has extensive modernization projects currently.

> Raise wages for once to attract talent (it has almost been a decade without wage growth, very bad for morale). FYI: the government shutdown threats (ie: federal employees / contractors don't get paid randomly) aren't helping.

This is hard to be fair. If you want to work in the USDS, you need to undergo a drug test (which is inconvenient even if you're not a drug user), an extensive security background check, and you can still be drummed out if they find something they don't like after you've already been working there for a bit (I asked). Compensation was fair though (~$132K/year, full medical benefits and TSP retirement plan access).

The tech industry is always going to be able to attract top talent better than government in this regard.

I agree with the rest of your post.

I think many are skeptical of the government getting into large-scale software development after the Obamacare website failures and the recent admission that the TSA spent millions on an iPad app that consisted of an arrow pointing left or right.

TSA paid $47,000 for the app, not millions. $47,000 may still be too much, but it's not millions. http://www.zdnet.com/article/tsa-contracted-ibm-47k-for-ipad...