Title: Auditing
1Auditing Statistical Sampling
2- Michigan Department of Treasury
- Roslyn Branner
- Mary Ann Nye
- Stan Weber
- Ryan
- Joseph J. Tomczyk
3Audit Goal
- Determine the correct amount of tax
- No More
- No Less
4Sampling Goal
- Create a subset of the data population that
mirrors the population activity - Acceptable to both parties
5Factors
- Transactional Data Availability
- Consistent Activity Throughout The Audit Period
- Acquisitions / Dispositions
- Records
- Number
- Complexity
- Adequacy
- Sufficiency
6Why Sample?
- Cost / Benefit
- Divert Professional / Compliance Staff How
Long? - Engage Independent Consultant How Extensively?
- Provide On-Site Resources to Auditor How Long?
7Considerations
- Audit Scope
- Plan / Agreement
- Data Validation
- Sample Evaluation
- Sample Projection
- Post-Audit Use of Results
8- Communicate Before
- Communicate During
- Communicate After
9Auditing Statistical Sampling Current Best
Practices
- Michigan Department of Treasury
- Tax Compliance Bureau - Audit
Presented by Roslyn Branner
10Why Do We Sample?
- Continuation of New Technologies
- Become more efficient in our Audit Processes
- A balance between accuracy and efficiency
- Cost and effectiveness
- Quality
11Statutory Authority
- To Sample
- None available at this time
- To Audit
- Revenue Act Provides the Authority to Audit
- Taxpayer is required by law to maintain records
for a four year statutory period. - Records stored by taxpayer are subject to review
regardless of the manner in which they are
maintained. -
-
12Statutory Authority
- Complete Records -- Mich. Comp. Laws Ann.
205.28(3) Mich. Admin. Code R205.4103 - Conduct Examination -- Mich. Comp. Laws Ann.
205.3(a) Mich. Admin. Code R205.4103 - Failure or refusal to file a return or pay the
tax, or the Department's belief that a return
does not give enough information to accurately
determine tax due, triggers the Department's
right to examine records, books, and so on, and
to audit accounts so as to assess tax. Mich.
Comp. Laws Ann. 205.21 . -
13Statutory Authority
- Subpoena Authority -- Mich. Comp. Laws Ann.
205.3(a) - Multi-State Cooperation -- Mich. Comp. Laws Ann.
205.28(1)(d)
14Records
- Records Documents, books, electronic data,
work papers, and/or any other information
relevant to determine tax due, tax reported or
tax collected.
15Current Staffing Structure
- All auditors have been trained on statistical
sampling procedures and how to use the software
(ACL). - Each auditor performs all steps required to
complete a statistical sampling audit. - Each audit goes through at least two levels of
management review.
16Sampling Methods
- Non-Statistical Sampling
- Manual
- Block or Judgmental Sampling
- Simple Random Sampling
- Statistical Sampling
- Electronic
- Stratified Random Sampling
17Statistical Sampling
- Pros
- Improves Accuracy
- More Efficient
- Virtually eliminates Taxpayer/Auditor Bias
- Cost / Benefit
- Performed in accordance with Generally Accepted
Auditing Standards - Reduces data entry errors
- Equal chance for all items
- Improved Quality
- Cons
- Inherent Sampling Risk
- Data for specified audit period is unavailable
- Repeated request for data due to auditor /
taxpayer miscommunication - Taxpayer or its IT Department is resistant to
providing the requested data - Not an exact science
18Sampling Risk
- Is assumed by all parties involved
- Can occur when the sample does not reveal the
true attributes of the population - The results of which can favor the Taxpayer or
the Department or both. - Minimized with proper planning
- Communication is a key factor
19Pre-Audit Planning Considerations
- Type of business
- Multiple locations
- Accounting Systems
- Internal Controls
- Decentralized or centralized accounting
- Changes in business practices over time
- Others
20Acceptable Data File Formats
- Excel
- Access database
- Dbase
- Delimited Text (Tab)
- Print Image (Report)
- Others XML, ASCII, EBCDIC, etc.
21 Media Types to Receive Data
- CDs
- DVDs
- Flash Drives
- External Hard Drive
- Email (not encouraged by The Department)
22Handling of Negative Transactions
- Ideally, all negatives would be matched with its
corresponding positive transactions and removed
from the sampling frame. - Remaining (I.e. unmatched) negatives are removed
and given to taxpayer for review. - Remaining negatives will be used to offset any
exceptions (I.e. errors) found in the sample.
23Handling of Negative Transactions
- Statistically, leaving the negatives in the
sampling frame distorts the desired confidence
and precision levels.
24Handling Missing Items
- Meaning
- A transaction where taxability can not be
determined without reviewing the source
document(s), however the source document(s) are
unavailable and cannot be provided. - Are considered an exception (i.e. taxable)
- Can not be replaced or substituted
25Handling Duplicate Transactions
- Should be identified and eliminated from the
sampling frame prior to pulling the sample. - Similar characteristics within the data should be
evident for the auditor to identify the
duplicates if theyre unknown previously. - Control total verification may reveal duplicates.
26Extraordinary Items
- Generally, no item is considered an extraordinary
item. - All items represent the activity in the normal
course of business - If Identified, can be pulled out of the
population prior to drawing the sample - If not identified before the sample is drawn,
will remain in the sample as all other items
27Factors in Determining Sample Size
- ACL software
- A customized module used in conjunction with ACL
which automates the sampling process. - Sampling Criteria
- Confidence Level
- Relative Error Tolerated or Relative Precision
- Minimal Sample items in each stratum
- Samples can be reproduced without the module
28Factors in Determining Sample Size
- Optimal Strata Break Criteria
- High Dollar Value
- Low Dollar Value
- Value of Pre-stratification Intervals
- Number of Optimal Strata Breaks
- Minimal Number of Items to be Selected
- Cumulative Square Root of the Frequency
29High Dollar Cut-off
- Includes only the higher transaction amounts.
- Detailed review only
- Will not be projected to other years, when all
data is available for the entire audit period. - Determined by auditor and agreed to by both
auditor and taxpayer.
30Low Dollar Cut-off
- Referred to as the De-minimus area
- Includes the lowest transaction amounts that are
greater than zero. - Determined by auditor and agreed to by both
auditor and taxpayer - Taxability is determined by projecting the
adjacent stratums percentage of error.
31Low Dollar Cut-off
- Purpose is to improve efficiency
- Taxpayer does not spend valuable time pulling
documents - Auditor does not spend time reviewing documents
- Not to be assumed as an area that will be
ignored. - Alternative methods
32Reviewing the Sample
- Review source documents to determine taxability.
- Opportunity for the taxpayer to review and
provide additional information not previously
made available. - Opportunity for the taxpayer to review remaining
negatives to off-set any determined exceptions.
33Projection Method
- Ratio Estimation
- Basic Formula
- Sample Error in s
- ------------------------- X Population Base in
s - Sample Base in s
34Evaluation Method
- Performed by a Statistician
- Performed on the largest statistical sampling
audits
35Performing a Sales Tax Audit
- Detail review of transactions in an electronic
environment - Statistical Sample
- Non-Statistical Random Sample
- Block Sample
36Performing a Use Tax Audit
- Statistical Sample of Accounts of Interest
- Non-Statistical Random Sample
- Block Sample
- Detail Review
37The Departments Responsibility to the Taxpayer
- Perform Statistical Sampling Audits in accordance
with Departmental policies and procedures. - Maintain communication with the Taxpayer
throughout the entire audit process. - Provided supportive documentation that clearly
communicate the results. - Maintain complete confidentiality of taxpayers
data and information.
38The Taxpayers Responsibility to the Department
- Full cooperation from the taxpayer.
- Provide records in a reasonable time frame.
- Maintain records according to federal and state
provisions. - Provide control totals
- Provide a proper Audit Trail.
39- The previous content was intended for
informational purposes only and is not to be
interpreted as official statements of the
Michigan Department of Treasury. The content,
and/or statements made during the presentation,
are not to be construed as promulgated rules,
bulletins or rulings of the Department and are
subject to revision pursuant to the effect of
legislation, court decisions, regulations and
official statements of the Department.
40Sample Types
- Statistical use of statistical formulas
- To determine sample size
- To measure sampling risk
- Non-Statistical Can be anything else
- Requires professional judgment
- Generally, needs all parties consent
41Negative Transactions
- Generally increases the probability of selection.
- Michigans uses a conventional approach, but not
the only approach. - For example, Texas has a five step rule that they
use to determine - If a negative transaction is going to be used in
the sample. - How the negative transaction is going to be used
in the sample. - Another approach is to treat all remaining
negative transactions in a separate strata.
42Missing Items
- Michigans uses a conventional approach, but not
the only approach. - Some states select replacement (additional)
transactions at the time the sample is drawn to
be reviewed if the source document for the
original sample item is unavailable. - While the sample is drawn from a specific source
document (such as an invoice), there is often
alternative source information (such as a plant
tour, discussion with plant manager, third party
vendor, etc.) that can be a valid substitute in
making a taxable / exempt decision.
43Other Sampling Considerations
- Stores
- Utilities
- Multiple populations / multiple samples
44- Communicate Before
- Communicate During
- Communicate After
45Summary
- Michigan Department of Treasury has published its
Audit Sampling Manual in March 2007 - When looking at using a sample to conduct a
Michigan Tax Audit - Look at the Taxpayers and the Departments
expectations. - Participate in the sample planning.
- Understand how the results will be used to make
audit determinations. - The sample and subsequent results are designed to
reflect the Taxpayers activity.
46