|
Kana -> RE: CORE IRS Issue Distilled in Black and White (5/20/2013 1:45:37 PM)
|
quote:
ORIGINAL: cloudboy
quote:
Accept the IRS did wrong, and let's move on from there.
I would like to know how the IRS with a staff of 200 can make reasonable exempt determinations over 70,000 applications a year. That's 350 rulings per worker, per year.
They can't.
You pull selective samples and extrapolate the results to the whole.
There's two ways auditors generally pick samples-judgmental or statistical.
Statistical is pretty straightforward-a computer random generates a certain amount of applications using a predetermined algorithm to select a number of applications to be audited. These algorithms tend to have sub-programs that ensure that a reasonable number are taken from each class to derive a good idea of how things are going with the entirety of the sampling universe.
This is what the IRS should have done. They have the computers. They have the programs. Hell, they've kicked out billions to develop these freaking things. It's the 2nd best way to cover your ass. The best way is simply to select 51% or more of the applications-then you can say you tested a majority of the universe and that the confidence ratio on the audit opinion is reduced to a minimal error rate on which the auditor can make a sound opinion backed with fact.
Judgmental is exactly what it sounds like. The auditor uses some sort of sorting technique to select samples. This is usually stratified-take X amount of applications from all stratas,(Think organization size, funding, location etc...), then audit them. Otherwise, the auditor uses some sort of sampling methodology to make their selections.
But, and this is the mammoth but here...
When an auditor does judgmental sampling, the auditor has to justify their sampling selection technique in the audit working papers. As in, has to. It's Generally Accepted Accounting Principles required, Generally Accepted Government Accounting Standards required, and thus, it's FREAKING FEDERAL LAW!
And they can't just throw out some wing nut justification-it's gotta be good. As in good enough to stand in a court of law.
These agencies know they get audited (Again, it's GAGAS-all audit agencies themselves have to be regularity audited or they aren't GAGAS compliant)by the IG, the supervisors know, the regional managers know-it goes all the way up the line. And when audits are written up, they are done with at least 1/2 an eye towards passing the future IG audit that's sure to come.In addition, IRS standards tend to be even more scrupulous than many other Fed agencies because they end up being litigated so frequently.
So, as common practice, these people should have been covering their ass ten ways to a hundred when they wrote up their sampling techniques. especially on a topic which was certain to be a political football.
Which is where we get to the critical point-no government audit is a one man show. They go throw multiple layers of review, starting with the working auditors on the audit team, then the team lead,a peer review, then the Supervisory Auditor, then the branch manager, a quality control person and most likely a Regional person is gonna look it over too.
So this wasn't a one man thing. It certainly wasn't a few out of control low level employees. It had supervision coming down from on high from the get go.
How can I be sure?
Because the first stage of an audit is where one decides how much needs to be sampled and the criteria for such. Nothing happens until that's done. no benchmarking, no timelines, and thus, no budget. And those things, they get supervisor approved.
As in always.
So let me repeat-an audit can't even begin until at least one layer of supervisors approved of the selection criteria.
Just saying.
|
|
|
|
