What Is Sampling and Testing in a Sales Tax Audit?
A lot of the goals of the first meeting with the auditor are to go through the process and try to keep the auditor on pace.
Very early in the audit, the auditor has to make a decision on the method of testing that they are going to use. In most situations, you want them to stick to the direct method of testing.
The auditor is either going to do the direct method of testing, which is based on those primary source documents, or they are going to use an indirect method of testing, which is a statistical sample.
The decision of whether the auditor is going to use the direct method or indirect method of testing can make or break your audit. Depending on which method they use, the auditor will run their test and will draft a preliminary audit report.
After that preliminary audit report gets drafted, the auditor and the representative will work to clarify any discrepancies and then they will produce a final audit report.
If the taxpayer agrees with the audit report, the audit is over. If the taxpayer disagrees with the audit report, they can appeal it.
Direct Method of Testing in California Sales Tax Audits
The direct method of testing is looking at source documents in the course of an audit and making sure that everything matches. That is the preferential way of dealing with sales tax audits because it is the most reliable; it is based on actual data and not statistical guessing.
To illustrate the direct method of testing, take a restaurant, for example. What a direct method testing would involve for a restaurant is looking at the bank statements for a given year, sales tax returns that were filed in that given year, 1099-Ks, merchant account processing statements for credit cards, internal accounting that was done during that year, and then the POS system reports.
You take those five or six pieces of data and you compare them across each other to make sure everything lines up.
For example, one of the most frequent errors that we see in sales tax audits with people who are under-reporting their sales tax is their federal income tax returns and their sales tax returns do not match.
To the extent possible, and particularly when the client has not done anything wrong, we want to keep everything to a direct method of testing. If the direct method of testing holds up and the auditor's unable to challenge it, then there is no real reason to go towards any statistical analysis.
The quickest and easiest way to wrap up an audit is to say, "Here are six pieces of paper that prove I do not have any sales tax liability," and going from there. The problems arise when you find discrepancies in those audits and then you have to move to an indirect method of testing.
Indirect Method of Testing in California Sales Tax Audits
The direct method of testing is very straightforward and involves testing actual source documents, lining them up and comparing them. When you have a breakdown in the direct method, then sales tax auditors will resort to what they call indirect methods of testing.
Indirect methods of testing is a fancy way of saying, "We are going to play guessing games with statistics." One of the indirect methods of testing is to audit past sales. The auditor will look at current sales and they will perform statistical comparisons between past sales and current sales.
The problem with statistics is you are taking a population of transactions and you are taking a sample of that population.
For example, say I have a restaurant and the restaurant had three years of sales that are included in the sales tax audit. Depending on which days I pick for my sample, it could significantly skew the results of the audit. If I pick weekdays, I am probably going to see a lower transaction amount, and I am probably going to see more cash sales.
On weekends where there are more large groups and more people go out to dinner, and there are larger transactions, I might see fewer cash transactions. I might see larger than average daily sales.
Two weekdays versus two weekends, if taken as a sample and applied to that population could yield very different results from a statistical standpoint. Even if the restaurant is diligent enough to track its daily sales, if I am using a statistical method to arrive at those sales, versus what the restaurant has, I could get a huge discrepancy.
One of the easiest tests the auditors run is through an observation test. The CDTFA will send an auditor into a business for a couple of days to look at the sales transactions, whether the employees are ringing everything up correctly and whether they are charging tax.
Then, the auditor will sit there and literally record every single transaction, and they will compare that against the POS system reports to see if there are any discrepancies. That is the observation test. If there is an error within that test, then they will perform certain actions based on that error.
The other thing they can do is they can take the POS system reports in a current period and look at the cash to credit card ratio. A lot of the time, they will take your credit card ratio, which can be verified by your 1099-Ks and say, "Okay, here is the percentage of credit card sales based on current. Here is the amount of cash sales and we are just going to project that across the quarter."
The problem that we see with indirect methods of testing, taking the cash to credit card ratio, for example, the auditor comes in and says, "I am going to do two days of testing in a business. I am going to come in on Tuesday at lunchtime and I am going to come in Sunday afternoon."
What the auditor is saying is that two days of testing are going to be used as a sample. They are going to take that sample and they are going to apply it to a population, with the population being a total dataset.
They are going to take a two-day sample and they are going to apply it to a three-year period of a population.
The three-year population is over one thousand days. The auditor is going to take two days of sales, and they are going to say, "These two days of sales, Tuesday at lunchtime and Sunday afternoon, are representative of this entire thousand days of sales." Hopefully, you can see the problems with that.
Some businesses do more credit card sales on the weekends and some businesses are seasonal and will have more frequent sales at certain times than others. With restaurants, families go out for dinner on Saturday nights and those tend to be larger checks. Larger checks tend to be paid by credit cards.
Little hinges swing big doors, and this is why indirect methods of testing are so dangerous. These little statistical changes can have a huge impact on the liability.
One of the reasons I advocate for having someone represent you in a sales tax audit who understands statistics, is because of these situations. We find so many statistical errors.
The auditors are not statisticians, so we find many errors in the way that they are doing their statistical analysis, the tightness of their controls, their procedures and the way they are applying samples to populations.
Whenever possible, you want to avoid indirect methods of testing and focus on direct methods because direct methods of testing are much more reliable than their indirect brethren.
Want To Save This Guide For Later?
No problem! Just enter your email address and we'll send you the PDF of this guide for free.
Different Types of Statistical Testing in California Sales Tax Audits
Audit Based on Verification or Accumulation of Taxable Differences
Auditors may decide to use verification of taxable differences rather than base the audit on total sales and claimed deduction basis. This method is preferred when records are available, but the verification of gross receipts or deductions is unnecessary due to the low number of transactions.
It may also be used when the taxpayer reported taxable measure is based on a listing of transactions, the capitalization of tax reimbursements, or by markup of taxable purchases.
A comparison between recorded and gross sales can disclose a number of discrepancies:
- Sales, gross receipts, fuel use
- A month in which a department or branch was not included in reported totals
- Classes of transactions or use that are erroneously considered non-taxable by the taxpayer
Audit Verification Methods
The accuracy of lists and tapes of taxable items is verified by determining if you failed to tax any items that should have been taxed and by verifying that all the items you charged tax on were indeed taxable.
Postings and computations are verified as well.
Auditors check your math on:
- Any taxes charged on customer invoices
- The posting of tax charges to a sales journal,
- Clerical accuracy of footings and posting of tax charges to a tax accrual account
- Your tax charge conversions to taxable measures.
Auditors employ several types of tests and sampling during a sales tax audit. The degree of complexity ranges from very simple (direct testing) to increasingly complex statistical sampling. There are also industry-specific tests, such as a markup test or pour test, which are frequently used in retail settings, restaurants and bars.
Taxable Measure Basis
An audit made on a taxable measure basis generally places emphasis on the verification or accumulation of taxable differences as compared to an audit performed on a total sales and claimed deduction basis using individual lead schedules. The auditor will verify that all sources of revenue and deductions have been examined.
Taxable measure basis may be preferable for the CTFA in a number of cases.
For example, when records are available, but verification of total gross receipts and deductions is not necessary because taxable transactions are few in number and the taxpayer has reported taxable measure only based on a listing of these transactions, capitalizing tax reimbursement, or by markup of taxable purchases.
Another circumstance is where the total gross reported is not an important factor in determining taxable measure.
For example, this is the case with service enterprises, contractors, public utilities, manufacturers and wholesalers (applicable for sales tax purposes).
Other cases include circumstances when direct audit of records will not yield results for the CTFA and indirect audit is necessary, or where the taxpayer has prepared returns on a taxable sales basis and audit time can be conserved by conforming to this method.
For example, where a grocer has used a method other than the “grocers method” to arrive at the taxable measure or the overall markup, as reflected by recorded gross sales, is unfavorable, the audit might be conducted in a more efficient manner through a verification of taxable sales.
The CDTFA makes a comparison between recorded and reported total gross an important procedure, for it may disclose that sales, gross receipts, or fuel used, for one month or one department or branch of a business was not included in the reported totals.
Through this comparison the CDTFA may discover classes of transactions or use erroneously considered nontaxable by the taxpayer.
If the taxpayer has reported on the basis of lists or tapes of taxable items, the auditor will verify the correctness of these lists by:
- Verifying that the lists include all items regarded as taxable by the taxpayer
- Determining if there were any items subject to tax not so regarded by the taxpayer.
If the taxpayer reported on a basis of the tax actually charged to the customer and has credited that amount, the clerical accuracy of the posting to that account, as well as the computations made in converting the tax accrued to taxable is measured by the following tests:
- The computations of tax charged on customer’s invoices.
- The posting of tax charges to the sales journal or other record where such charges are summarized.
- The clerical accuracy of the footings and the posting of the tax charges to the tax accrual account.
- The mathematics of the conversion of tax charged to taxable measure reported
For example, debits to the accrual account will be scrutinized by the CDTFA to determine that these charges represent proper deductions from the amount of tax accrued.
In addition, the auditor will determine the effect on the accrual account of allowable bad debts, tax-paid purchases resold and other adjustments which legitimately reduce the taxpayer’s accrued tax liability. A reconciliation of the tax accrual account and the tax reported will be made by the CDFA auditor.
Consideration also will be given to sales and use taxes collected for and paid to other states. Deductions claimed or netted will be tested by the auditor to ensure they are allowable.
The California Tax Attorney’s Explanation to Short Tests and Sampling
The CDTFA uses short tests to come up with a decision as to whether to proceed or to accept as correct that item being tested. In fact, the CDTFA encourages its auditors to use short tests when a taxpayer’s records are in order.
However, at any time, a short test can be expanded into full examination.
A short test may be defined as the examination of any record, supplemental data, original detail, etc., for any purpose. A short test audit may be a combination of several short tests.
A short test might be, for example, the review of an income tax return to see if the markup over cost is acceptable for the type of business, a spot check of sales invoices for proper tax accrual, etc.
The nature of the short test places a great deal of emphasis on the individual auditor’s judgment. The auditor evaluates the results of a short test and comes to a decision as to what to do.
In certain businesses where the number of transactions is large (e.g., department stores), the CDFA auditor may design a controlled test for a short period. An example might be the operations of one day or less.
This test, even though formal in nature versus spot checking, would be construed as a short test because this forms the basis of a decision to stop testing or to proceed.
If the auditor decides to proceed, this original test might be the nucleus of an expanded audit program.
Size of test period. Auditors use their experience and exercise judgment in determining the size of test periods. The following principles should be considered in selecting a test period:
- The size should be adequate to insure reasonable accuracy
- The auditing time required should not be excessive in relation to the problem
There are two broad categories of testing used in sales and use tax auditing – statistical sampling and block sampling. The second category of testing is used when statistical sampling cannot be used.
The auditor will examine any record, supporting data, or other information to determine whether or not to proceed with the audit on that document or if it can be accepted as correct without further examination or investigation.
A short test can be expanded into a full examination any time but saves time when a full examination is clearly not required to establish accuracy. In some instances, a short test is designed as a control test of documents over a short period to determine the most representative sample size for audit.
Testing is split into two categories: statistical and block sampling. Block sampling is used when statistical sampling is not possible.
The auditor also looks into the conditions for employing the tests, especially when looking at the consistency of units sold or in business characteristics during the test period.
The size of the test period depends on what it takes to provide reasonable accuracy without excessive effort in comparison to the problem. Test periods are usually complete months or quarters. If daily or weekly controls are established, periods shorter than one month may be selected.
If your business has excellent controls, the test period will probably be a very small sample of the audit period.
The CDTFA looks for a number of conditions to employ tests – most important of which is consistency of units sold, consistency in business characteristics in the test period and so on.
For example, it may be that what is being tested is a hodgepodge of various deductions. In this case, the base would be all the deductions recorded and claimed.
Verification is then made by the auditor in accordance with the taxpayer’s method of computing the claimed deductions.
In general, when auditing a business with good internal controls, and a good accounting system, the test period may be a relatively small portion of the total audit period.
However, in an audit of a business with little or no internal controls, the test period most likely will cover a larger proportion of the audit period.
If records are available, the periods selected for testing will be spread over the entire audit period so that samples can be taken of all years and all seasons of the year.
The size of each test period, in addition to the above considerations, will depend on the number of documents required to be examined.
Usually, the test periods consist of complete months or quarters, but periods of less than one month may be selected by the auditor if daily or weekly controls can be established.
CDTFA statisticians have established that several short tests over the audit period are superior to one equivalent long period. For example, a test of three months scattered throughout the audit period will give better results than a one-quarter test.
The CDTFA checks for errors and statistically applies them to the general number of transactions. If errors are irregular and non-recurring, the CDTFA will eliminate them from the test and be excluded from the calculations of a percentage of error.
For example, any sale that is rare or out-of-the-ordinary, can be considered, for classification, as nonrecurring.
An audit made on a test basis where there is no supporting detail (i.e., no detailed journals) is conducted similarly to a detailed audit where there is no support for the claimed amount.
However, the taxpayer will be requested to prepare supporting schedules for the test periods only.
The proposed measure resulting from the projection of the sample results will be compared and analyzed for reasonableness by looking at the taxpayer’s business as a whole.
If the results appear unreasonable, the auditor will discuss the situation with the taxpayer. The auditor and the taxpayer will need to come to some kind of agreement as to whether or not the results are representative of the business for the time period in question.
The auditor will, whenever possible, discuss the use of test periods with the taxpayer, include the taxpayer in the development or selection of a sampling plan, and endeavor to obtain a concurrence.
In fact, very often during the audit, the auditor will discuss and consult with the taxpayer about the audit’s procedures and techniques, to ascertain necessary information.
If, during the course of a sample, a document cannot be located, normal auditing procedure requires the auditor to ascertain the reason for the missing or incomplete documents.
When the investigation fails to reveal any specific reason, the auditor may first determine whether there is any acceptable alternative evidence.
The auditor and taxpayer will work together to obtain missing documents or the auditor can ask for additional documents that may indirectly provide information sought by the auditor.
The auditor will develop a sampling plan to outline methods of testing, the timeframe and so on.
Prior to determining the type of testing to be used in a given audit situation, the auditor must make a thorough examination of the business operations for the period under audit.
This examination includes a review of source documents, changes in business activity, and changes in accounting procedures and key personnel.
A plan is usually completed with assistance and input from the taxpayer. The information and methods documented in this plan are not binding on either the taxpayer or CDFA staff.
The sampling plan can and will be continually evaluated (and changed, if necessary) based upon information obtained during the audit process.
However, if any deviation from this sampling plan is required, the deviations are fully explained and discussed with the taxpayer.
CDTFA staff must first try to obtain any data or documents which should have been retained in accordance with Revenue and Taxation Code section 7053 from the taxpayer.
However, if all other available avenues of information have been exhausted and approval of the district administrator has been obtained, CDFA staff may request the information directly from the taxpayer’s financial institution either by obtaining the taxpayer’s authorization or by issuing a subpoena duces tecum.
Procedures for requesting records directly from a financial institution must comply with the California Right to Financial Privacy Act.
CDTFA Audits: Prior Audit Percentages of Error (PAPE) Program and Cut-Off Techniques
The Prior Audit Percentages of Error (PAPE) Program involves the use, under certain circumstances, of a percentage of error developed in a prior audit for the sales or accounts payable portion of a current audit. It can be a valuable tool in streamlining the audit process and is designed to reduce the time it takes to complete an audit and minimize the burden on taxpayers.
When planning the audit, CDTFA supervisors and auditors evaluate whether the taxpayer is eligible for the use of a PAPE. This evaluation is conducted whether or not the taxpayer has already requested the use of a PAPE. If the taxpayer is eligible for the use of a PAPE, the auditor discusses the PAPE with the taxpayer as soon as possible rather than wait for the taxpayer to request it.
To qualify for the PAPE, the taxpayer must have at least one prior audit and must meet the number of conditions. One of the conditions is consistency of business operations during prior audit and current audit.
Minor changes are generally ignored. It is important to remember that the use of a PAPE is limited to the current audit period as a PAPE cannot be used in two subsequent audits.
The CDTFA also uses cut-off techniques. “Cut-Off” is that point in the audit program where the auditor has accumulated sufficient data to support a reasonable conclusion or opinion based on acceptable audit standards. It might be defined as when to stop testing or examining data.
It could be used for whole examination or for a specific task or test and the auditor has discretion in this case. When a prospective cut-off point is reached, a decision is made by the auditor whether to accept the test results, alter the audit approach, or discontinue the audit.
Basically, an important consideration here is given to existence of errors. It is up to the auditor to decide if errors exist, whether to keep testing or disregard errors.
The auditor can decide to accept test results and the cut-off point is reached then. The auditor considers materiality of error, frequency of error, and other factors.
The principle of whole dollar auditing (i.e., dropping cents) is used by the CDTFA as a time-saving technique.
In whole-dollar auditing, cents are eliminated at the earliest practical stage in an accounting sequence and only whole-dollar amounts are recorded thereafter. In this case, any $.49 cents and below are rounded down, and $.50 cents and more are rounded up.
A whole dollar audit is not used when the taxpayer objects (the auditor may still try to convince the taxpayer by explaining this method), when computing markup form shelf orders or when precise data is necessary.
During an audit, the auditor will examine multiple records to find undisclosed and unreported taxable transactions. The general ledger accounts will be examined for debits and credits.
The auditor may examine general journal entries, correspondence, contracts, sales or revenue invoices, cash receipts records, accounts receivable ledger, partners’ drawing accounts and employees’ advance accounts, purchase journal for sales at cost or returned merchandise, including inventory accounts.
The auditor will try to reconcile sales or revenue reported to general ledger. In cases of missing or flawed records, the CDFA auditor will do gross profit and net worth analysis. Auditor will review tax returns as well.
Overall, the CDTFA auditor will make a significant effort to ensure a good relationship with the taxpayer and will try to make the audit convenient for the taxpayer, as long as the taxpayer does not create unnecessary obstacles for an audit.
Nevertheless, it is advisable for the taxpayer to cooperate with the auditor to ensure a smooth and efficient auditing process. It is also really helpful to ensure that all books and records are always kept in order, even if the taxpayer does not expect an audit
When the CDTFA cannot verify through direct audit, it will perform a markup audit. A markup is the amount added to the cost of a good or service to obtain sales tax from the buyer. It is a percentage of the gross profits divided by the cost of goods sold.
A markup audit is often used to prove materially misstated or fraudulent accounting, especially for cash-intensive businesses.
Auditors are trained to spot illegal programs that periodically make a certain amount of sales and cost of goods disappear from a point of sale system (POS) while simultaneously informing an owner how much cash to skim illegally from the POS to make things even.
Auditors will also approach a vendor to compare whether the dollar amount of goods sold to the business under audit matches what the business’s cost of goods sold account for the same period.
However, in the context of a sales audit, a markup test is kind of what it sounds like. It determines the markup in the aggregate of taxable products that are being sold.
Take a restaurant, for example, the auditor will look at all the menu items, and they will have somebody break out food costs and say, "Okay, if you are selling burritos, tell me all the ingredients that go into the burrito and give me an estimate of how much it cost you to make that, and then let us see what your markup is."
Markup tests are really dangerous because markup tests generally vary across different products.
A restaurant selling beer can have a certain markup. A restaurant that is serving shots coming out of a bottle is going to have a higher markup on liquor than with beer. In trying to take an average of those two things, it is really difficult to get an appropriate average markup.
We try to avoid markup tests whenever possible. If we have to go through them, we will try to limit the scope of the markup test to make it as easy as possible on the client and to make sure that the results that we are getting are really consistent
Markup tests are very common, particularly in retail settings, with restaurants, and with a variety of other businesses. They really should be avoided at all costs.