VII. The Audit of VAT
- Alan Tait
- Published Date:
- June 1991
Milka Casanegra de Jantscher, Carlos Silvani, and Graham Holland
If VAT auditing is not reasonably effective, VAT administration will be ineffective. Tax administration may be thought of as a series of links making up a chain. The strength of the weakest link determines the strength of the chain as a whole. When tax audit is weak, tax administration will also be weak. If other enforcement systems such as the identification of nonfilers and the control of delinquent taxpayers (the collection function) work effectively, tax noncompliance will take the form of evasion when auditing is weak.
To minimize the gap between the tax reported by taxpayers and the statutory tax, an adequate audit plan should be implemented. This plan should provide for broad audit coverage of taxpayers, including appropriate coverage of different economic activities and taxpayers classified according to amount of sales. It should include different procedures to fight each form of noncompliance in the most appropriate manner. In addition, there should be programs to prevent noncompliance.
Typical Forms of Tax Noncompliance
There must be clear guidelines to deter tax non-compliance and a specific amount of resources should be allocated to combat each type of tax noncompliance. The most typical forms of tax noncompliance are discussed below.
Tax defrauders are taxpayers who engage in fraud, such as altering or counterfeiting documents. Thus, their actions go beyond tax violations and fall into the realm of crime. Fraud should be punished even if there is no loss of revenue. To combat fraud, special programs to audit a small number of cases should be carried out. Taxpayers audited should be those for whom preliminary checks reveal the possibility of fraud. Harsh penalties, including jail sentences, should apply. The penalty should also be made public, since the primary objective is to punish the offender as an example to others.
A different strategy should be used to deal with taxpayers who underreport the tax due but do not commit fraud. The purpose is to dissuade the violator so that he will change his behavior and not repeat the offense.
The level of evasion depends mainly on the probability of detection and the severity of the penalty. The way to increase the probability of detecting and punishing VAT evaders (i.e., increasing the risk associated with evasion) is to develop a strategy based on broad and swift audit. The errant taxpayer should be audited for specific items of noncompliance, without investigating whether he has committed other violations. Thus, the VAT audit coverage (the ratio of audits to taxpayers) will be high and, therefore, so will be the risk associated with evasion.
Priority should be given to increasing the audit coverage and the probability of “effectively” sanctioning tax evaders. Effectively, in this case, means in such a way that the taxpayer actually pays a penalty and is not just handed an assessment that will never be paid in full.
Another fact must be considered. If the probability of penalty is low, the taxpayer will accept the risk, even if the penalty itself is sizable. After all, taxpayers do travel by air knowing that they may have to pay the maximum penalty in case of an accident. They accept the risk because the probability that it will occur is low.
Tax avoidance is practiced by taxpayers who take advantage of loopholes in tax laws to underreport their tax, that is, those who take advantage of circumstances that are not clearly defined in the law or can bear different interpretations. Special audit programs must be implemented to detect tax avoidance. Once a particular form of avoidance is detected, the problem can be countered by modifying the tax statutes and their interpretation.
Besides allocating resources to combat tax non-compliance, preventive programs should be carried out to convince taxpayers that evasion is bad business for them. To this end, the tax administration should show taxpayers that it has both the information and the operational capacity to impose sanctions effectively should they engage in evasion.
Preventive audit programs should be drawn up to let taxpayers know about the information available to the tax administration regarding certain transactions, such as imports, purchases, and sales. In this case, the purpose of the auditor’s visit, or of a notice sent by mail, is to make clear to the taxpayer that the administration has data that will be compared with the returns received to detect any inconsistencies. Obviously, it is not necessary to inform taxpayers of every detail regarding the information which the tax administration has or does not have. What is important is to demonstrate to a significant number of VAT taxpayers, before they engage in evasion, that the tax administration has data on business transactions and that this information will be used to detect evasion.
Types of VAT Audit
The terminology of tax administration varies widely around the world, and it may be useful to clarify what is and what is not meant by the term VAT audit. In the course of its compliance work, the VAT administration comes into contact with a taxpayer in a number of situations.
The Advisory Visit
This contact should be regarded primarily as a part of taxpayer education. At the stage when a VAT is introduced, this visit takes place as soon as possible after the taxpayer has registered, and preferably before the VAT implementation date. When the VAT is in full operation, this visit would take place as soon as possible after a new taxpayer has commenced business, and may well be associated with other start-up activities of the tax or VAT department (e.g., possibly pay-as-you-earn (PAYE) advisory or income tax advisory visits).
As mentioned in Chapter III, this initial visit serves two purposes: (1) to educate the taxpayer about his obligations and to ensure that his records, etc., are adequate; (2) to gather some basic information that may help the administration in its future dealings with the taxpayer. This visit is not regarded as a VAT audit. Copies of an advisory visit report and checklist forms used by one VAT office are reproduced in Annex I.
Some administrations describe a desk or an office audit as checking the return for arithmetic accuracy, calculating ratios from the return data and checking for completeness, and so on. It is conducted in the tax or VAT office without discussion with the taxpayer. This activity is essentially clerical and is best automated as far as possible, for example, by having all arithmetic checks and comparisons made within the computer system when the return is processed.
This type of audit, also known as a verification audit, control audit, or routine control audit, is the main subject of this chapter. It differs from the income tax audit in that it is usually less complex and less extensive. It is generally a check of selected aspects (see below for details) of the taxpayers’ records; its goal is to minimize evasion through frequent control checks so that evasion can be quickly identified and penalized.
This refers to the in-depth investigation of taxpayers generally conducted after a VAT audit has indicated the likelihood of fraud. It is similar to an income tax fraud investigation, and there are good reasons for combining the two activities, since a taxpayer committing fraud in one tax will almost certainly be evading the other.
Main Features of a VAT Audit
The main features of successful VAT audit systems are as follows:
(1) Taxpayers are selected for audit on the probability that they will be evading tax.
(2) The frequency of a VAT audit is far greater than is customary for income tax.
(3) The scope of a VAT audit is quite different from that of an income tax audit. The VAT audits consist primarily of short, verification visits to detect evasion. They are not intended to be a complete financial audit, duplicating or superseding the income tax audit.
(4) Fraud investigations are conducted by specially trained auditors.
Chart 1 traces the main steps in a VAT audit. It illustrates the flow of enquiry that an effective audit could take, including the cross-matching of data in the return against the following materials:
Chart 1.VAT: Audit Model
(1) Financial records such as invoices, work sheets, and bankings. Clearly there should be a very close relationship between the total value of invoices issued, the periodic summary of sales in a VAT work sheet, bankings, and the output tax data shown in the VAT return. The checks should confirm that there are no missing invoices and that sales have been properly classified into taxable, exempt, and zero rated. If there is a material discrepancy between any of these, the reason should be established by discussion with the taxpayer, by making further checks of the business records, and, if necessary, by interviewing suppliers or customers. An unexplained discrepancy between any of the above records may be the result of innocent error or fraud.
Similar matching will be required in relation to credits claimed for inputs. For this, the comparison will be between the invoices held for a specific period, the work sheet summaries, checks written, records of cash expenses, and credits claimed in the return.
(2) External information such as customs documentation and documentation held by customers and suppliers. Information gathered from the customs department can assist in the independent verification of the data in a taxpayer’s return. Checks of bill of entry documents could help to identify excess input credit claims or possibly indicate an undisclosed business activity.
(3) Physical records including units of production, inventory, and material use data. A review of the physical aspects of the business could be made for two reasons. First, a quick check of the premises, nature of machines, etc., may indicate the types of purchases and sales that should be expected and may give some clue about the size of the business. More detailed calculations of estimated sales and purchases might be needed if there are no financial records or if an apparent discrepancy must be explained. The types of comparisons that can be made will vary widely among different kinds of businesses, but could include estimating sales, purchases, or value added on the basis of the production capacity of machinery, raw material and electricity usage, transport and cartage documents and costs, number of employees, salaries paid, interest paid, area of factory premises, and licensing data or production returns held by other government departments.
(4) Some administrations collect sample invoices from customers and suppliers. This procedure extracts details from (or copies) a small selection of invoices during an audit and posts them under the name of the other taxpayer to the transaction pending his audit. Invoices issued by a taxpayer can ultimately be checked to verify the input credit claims of the customer, while those received by a taxpayer should be reflected in his supplier’s output tax calculation. However, such sample invoices are only a minor part of the audit and could be classified as contributing to the information under “preventive action.”
(5) The filing of amended (corrected) returns in which the taxpayer admits his errors or omissions as detected by audit should be encouraged. Because they are filed by the taxpayer himself, amended returns are not subject to costly and time-consuming appeals procedures. Even if the taxpayer is unwilling to acknowledge the full amount of tax the administration says he owes, in some cases it is preferable to accept “voluntary” amended returns in which the taxpayer agrees to pay a portion of the tax due than to make an official assessment for the full tax that the administration believes was omitted. Official assessments generally give rise to appeals, so it is preferable to reach a settlement to avoid judicial proceedings. It should be emphasized that the primary purpose of tax administration is to collect taxes, not to engage in lawsuits. However, taxpayers should not be given the impression that the administration is always willing to compromise.
An essential difference between a VAT audit and an income tax investigation is in the time devoted to the activity. While an income tax audit or investigation may take several months to complete, the standard VAT audit should be completed in one or two days.
The VAT audit is a field activity conducted primarily at the taxpayer’s business premises—it cannot be done from the office.
The precise nature of the checks made during audit will vary, depending on the nature of the VAT law. However, the basic procedures will be almost the same and a typical audit is likely to involve the following:
(1) A check of all third-party data held in the office under the taxpayer’s name against his accounting records. This could include information culled from other audits, income tax or PAYE files, import data, or information collected from customers or associates. The data may also include copies of VAT invoices obtained from audits of other taxpayers.
(2) A review of the accounting records for the selected return period and/or other aspects of the business to ensure that they are adequate and to identify possible weaknesses for further examination. Internal control procedures require particular attention.
(3) Confirmation of the arithmetical accuracy of input and output tax calculations, tracing from the return back to work sheets and the accounting records.
(4) A sample check of source documents supporting both input and output tax calculations (bank statements, invoices, deposit slips, till tapes, etc.).
(5) A check for adjustments for cash and goods taken by proprietors and for the inclusion of cash sales.
(6) A comparison of external data held (such as payment schedules from customers) with books of account and returns.
(7) A check on exceptional items such as the preparation of self-supply invoices and second-hand goods credits.
(8) A check of invoices for input credits claimed on private purchases.
(9) Verification that exempt and zero-rated goods have been correctly accounted for.
(10) Checks should be made on deviations from the standard business operating ratios (e.g., input:output, exempt:taxable sales) and to establish any reasons for significant departures from industry standards.
(11) Sample checks can be made from purchase invoices to inventories, and the reasonableness of the latest inventory can be assessed in the light of recent purchase and sales data.
(12) Production records should be reviewed as an aid to assessing the validity of raw material input credits and output tax liability.
(13) Quantify any discrepancies, ensure that any evidence held and any explanations are recorded, assess whether there are any indications of fraud, and prepare report for supervisor.
The complexity of the task and the importance of properly recording the results so that violators can be penalized or prosecuted make it worthwhile to develop checklists and standard reporting formats. Examples of those adopted by the New Zealand authorities are reproduced in Annex II.
The types and level of evasion will vary with the complexity of the VAT adopted. Multiple rates and widespread exemptions and zero rating will increase the opportunities for evasion (and genuine errors) and also will give rise to types of fraud not found under simpler systems. Some instances of typical VAT frauds are given in Annex III.
To maximize the effect of audits, a case selection system should be developed, taking into account the following:
(1) The larger share of the audit should not be devoted to the audit of large taxpayers. Many countries overallocate resources for auditing large taxpayers, because the biggest differences between taxes due and taxes paid can be found in this group. Moreover, productivity (the additional tax assessment) per case audited is higher as well. Nevertheless, the total revenue obtained from all taxpayers may be lower if the audit of medium and small taxpayers is neglected. Large taxpayers are unlikely to engage in gross evasion, such as making uninvoiced sales or falsifying the value of their inventories (among other reasons, their own organization and accounting systems would not permit this). Therefore, total collections may increase more if auditing is directed to medium and small taxpayers, because of its positive effect on the compliance levels of taxpayers in general.
In some countries, audit staff is assigned almost exclusively to the audit of large taxpayers. The rationale is that since concentration of economic activity is very high, the audit effort should also be concentrated. If, for instance, 5 percent of taxpayers account for over 70 percent of tax revenue, then it is assumed that audit should concentrate on these 5 percent. (Table 1 shows the concentration of VAT collections in a selected European country.) Such a policy may lead to the unwanted result of increasing the concentration of collections among few taxpayers. In the extreme case, if the administration does not act on the remaining 95 percent of medium and small taxpayers, it is possible to arrive at a situation where 5 percent of taxpayers will account for 100 percent of tax revenues. It should be stressed that these comments pertain only to the audit of large taxpayers and not to other enforcement activities, such as monitoring nonfilers and delinquent taxpayers, where large taxpayers should receive priority.
|Number of Taxpayers Ranked by Size||Cumulative Number of Taxpayers||Percentage of Taxpayers over Total Taxpayers||Cumulative Percentage of Taxpayers||Percentage of Collection over Total VAT Collected||Cumulative Percentage of Collections|
Total VAT revenue has increased considerably in countries where the audit strategy has been changed to cover medium and small taxpayers. In those countries, despite a high degree of concentration of economic activity, the additional revenue has come mainly from medium and small taxpayers. As illustrated below, in Chile, for example, changes in audit strategy led to an increase of 55.5 percent in VAT collections during 1979-82. This rise was due to increases of 16.5 percent in collections from large taxpayers, 86.9 percent from medium taxpayers, and 185.6 percent from small taxpayers.
|Chile: Increase in VAT Collections, 1979-82|
|Taxpayer Size1||Percentage of Increase|
|Chile: Share of VAT Collections According to Taxpayer Size|
|Percentage of Collections|
Large taxpayers were considered to be those with annual sales over US$6,000,000; medium from US$600,000 to US$6,000,000; and small under US$600,000.
Large taxpayers were considered to be those with annual sales over US$6,000,000; medium from US$600,000 to US$6,000,000; and small under US$600,000.
It should be noted that the audit of the largest taxpayers may be beyond the capability of a typical VAT auditor. Audits of large enterprises need to be conducted by a high-level inspector who should be a qualified accountant with wide experience in auditing. The scarcity of such officers and the similarity of the task with the income tax audit of larger taxpayers, as well as the need to avoid duplication of effort, will generally mean that both should be conducted at the same time and by the same auditor. Therefore, there must be a balance of audit between large, medium, and small taxpayers. The following criteria will help to ensure that such a balance is achieved. In practice, it will be found that most audits will be of small- to medium-sized taxpayers.
(2) A Discriminant Function System1 to select taxpayers for audit will help maximize the effectiveness of audits. The DIF should consider such items as economic activity, sales volume, reported annual markup, and discrepancies resulting from cross-checking information systems.
(3) Ideally, there should also be a mechanism for ensuring that businesses that have not been audited within a reasonable period are selected. Otherwise, a business might escape selection for audit. The maximum length of time between audits will vary, depending on the nature and the size of the business: A maximum period of one year for the largest businesses, four years for medium, and seven years for small businesses might be appropriate. (In one country, the average period between audits was 33 years; clearly, taxpayer compliance was unlikely to be increased by this probability of audit. Even a country with a reputation for efficiency such as Sweden audited only 0.7 percent of VAT taxpayers in 1986.)
Audit Support Systems
Systems should be developed to support tax auditors in conducting field audits. For example, when a tax auditor visits a taxpayer he should have with him a computer printout containing data from the returns filed by the taxpayer. Some of the audit support systems that should be developed are described below.
VAT Summary Records
VAT summary records should be computer printed and contain the data furnished by the taxpayer on his return. Tax auditors should conduct the audit using the information on these records. To guide the auditors’ work, the computer should indicate on the summary records the existence of any errors, such as arithmetical errors. The records should also contain ratios such as sales/ purchases as well as the modal (or most frequent) values of these ratios for similar taxpayers (i.e., taxpayers with similar economic activity and turnover). Basically, the records should contain three kinds of data.
(1) Data reported by the taxpayer for various tax periods, such as reported sales by rate, reported purchases by rate, carryover of excess credits, and requested refunds. Any arithmetical errors found by the computer should also be shown here.
(2) Statistical information, such as modal value (most frequent) of markups reported by taxpayers similar to the taxpayer to be audited.
(3) Information collected from outside sources on the taxpayer’s transactions, including details on amount, date, and kind of activity, such as imports, sales to the taxpayer by his suppliers, and purchases from the taxpayer by his clients.
Purchases made by large public enterprises and government agencies should be checked to determine whether they have been reported as sales by the taxpayers who acted as suppliers. This type of evasion is common in some countries. Indeed, there have been cases where sales to the tax administration itself were not reported by the suppliers.
Purchases and Sales by Enterprises
Information on purchases made by medium-sized and large enterprises should be checked, for the same reason indicated above: to detect evasion by suppliers. On the other hand, it is also important to process data on sales made by medium-sized and large enterprises. In this case, the objective is to detect those purchasers who over a long period of time report total sales below the purchases they made from these enterprises.
Cross-checking information on purchases and sales should be carried out selectively, as is currently done in Bolivia and Chile, as well as in many other countries, by cross-posting a small sample of invoices collected at audit. Only information on medium-sized and large enterprises should be processed. Moreover, transactions should be selected according to such criteria as the minimum transaction size and the kind of merchandise or service. Efforts have been made to match each taxpayer’s sales invoices with the corresponding invoices in the records of the purchasers, which in theory, should provide a comprehensive record of trade and avoid false claims for input tax deductions. Such an undertaking, however, would require large computer resources and many staff. In practice, it has been found that it cannot be done effectively because of errors, incomplete returns, and the sheer magnitude of the task. In any case, cross-checking would not help detect unrecorded retail sales which can be major sources of evasion.
To detect sales not reported by importers, data on imports must be processed. By matching actual imports with the sales reported by taxpayers, it is possible to detect whether there are sales not reported in the VAT returns.
Payments by Enterprises to Others
When services are liable under the VAT, it is advisable to process data on payments made by enterprises to others, especially for taxable services received. Both in this case and in the others referred to above, priority should be given to processing data that enterprises can furnish to the tax administration on magnetic tapes, diskettes, etc. In this way, large volumes of information can be gathered at very low cost.
The use of the markup ratio to select VAT taxpayers for audit has proved particularly effective. The markup ratio is calculated for each taxpayer by dividing total sales reported for a given period, say one year, by total purchases in the same period. Priority for audit should be assigned to taxpayers who are reporting markup ratios below the modal value (most frequent value) reported by taxpayers within the same or a similar economic activity.
For illustrative purposes only, Table 2 shows the distribution by sales/purchases ratio of VAT taxpayers in a selected country. It can be seen that in a given year 28.91 percent of all VAT taxpayers were reporting annual sales equal to or less than their annual purchases (sales/purchases ratio equal to or less than 1.00), which means that they did not have a positive margin of sales over purchases with which to pay wages, financial outlays, and other expenses not subject to the VAT. In the wholesale and retail sector, the proportion of VAT taxpayers reporting annual sales equal to or smaller than their annual purchases amounted to 33.68 percent. Over 75 percent of taxpayers in this sector reported a markup equal to or less than 20 percent (sales/purchases ratio equal to or less than 1.20). A 20 percent markup is a reasonably low percentage, and most taxpayers should be above that level.
|Annual Sales/ Annual Purchases (Markup Ratio) From-To||All Economic Activity (In percent)||Cumulative Percentage||Wholesale and Retail (In percent)||Cumulative Percentage|
|6.00 or more||5.34||100.00||1.18||100.00|
Modal value (most frequent value).
Modal value (most frequent value).
The above comments are obviously only illustrative, and to define adequate ratios in a specific country would require a detailed review of each of the main economic sectors. It should be noted that large differences may exist even within a given economic sector. Within the retail trade sector, for example, markups vary substantially between gasoline stations and furniture stores, because of differences in inventory turnover. However, usually tax administrators in each country have a good “feel” for margins and their knowledge can be built into the DIE
It is worth noting that in the country used for illustration, after some years of broad audit coverage and using the markup ratio to select returns for audit, there was an increase in VAT revenues as a percentage of gross domestic product. Furthermore, the number of taxpayers reporting annual markups equal to or less than 20 percent decreased by about two thirds.
The Frequency of Audit
The frequency of audit refers to the average audit cycle which depends on (1) the size of the taxpayer base; (2) the number of audit staff; and (3) the average time taken to complete an audit.
An income tax audit or investigation can take considerable time because of its complexity, and it is not uncommon for audit rates to be as low as 1 percent or 2 percent of taxpayers a year and rare for the rate to exceed 10 percent. By contrast, VAT administrations find that better compliance can be achieved by more regular, short, and targeted audits. This is because the VAT payment cycle is shorter, requiring speedier identification of deficiencies (possibly a month for VAT as against a 12-month payment cycle for income tax), and also because it is a simpler system. The frequency of VAT audits on average tends to be higher than that for the income tax: for instance, it is once in 4 years in the United Kingdom and once in 5 years in New Zealand (although there are examples of very low average likelihoods of audit, such as, once in 30 years).
Management of Audit
The role of headquarters in audit management includes the following: establishing national audit targets; allocating audit budgets to regional and district offices; allocating financial and staff budgets for audit activity; monitoring the performance of national, regional, and district offices against budgets and targets; preparing and distributing schedules of files from which audits are to be selected; maintaining computer system to select cases and to record progress and results of audits; and supporting regional and district offices in their activities by providing training material, legal assistance, audit systems and forms, and other resources.
The tasks of local management are to allocate audits to staff according to their experience and skills; assist auditors with more complex investigations and assist at difficult interviews; monitor the progress of individual audits, with particular emphasis on the closing and extension of cases, and to check the quality of audits; identify training needs and where appropriate conduct the training of own staff; and review auditors’ reports for quality and completeness of audit.
VAT auditors do not need to be qualified accountants. However, they do need to have a basic knowledge of accounting, be fully trained in the interpretation of the VAT law, and be able to deal tactfully with taxpayers. The training requirements and numbers of staff needed are discussed in detail in Chapter V. As mentioned in that chapter, about 40 percent of the total VAT staff should be allocated to audit duties. Ideally, there should be considerable scope for a fairly long promotional ladder within audit, so that as staff are trained they can move to more complex audits and eventually to examinations of serious fraud.
Evaluation of Audit
The effectiveness of audit should not be measured in terms of its direct revenue yield—that is, the amount of additional tax assessed—but the extent to which audit contributes to better tax compliance. Judging the effectiveness of audit by the amount of additional tax assessments is similar to evaluating traffic authorities by the fines collected from drivers. The basic objective of tax administration is to maximize voluntary compliance, that is, to minimize noncompliance. In this regard, a highly visible revenue agent may be the most effective way of ensuring compliance with the law, even if it does not directly yield additional revenue.
The audit effort should be evaluated in terms of the quantity and quality of audits conducted and of “voluntary” collections generated, that is, the filing of amended (corrected) returns by taxpayers after an audit.
Annex I. Examples of VAT Advisory Visit Forms
Annex II. Examples of VAT Audit Forms
Annex III. Examples of Typical VAT Frauds
|Deposit to private bank account||Taxpayer admission|
|Till tapes do not reconcile to bankings|
|Missing sales invoices|
|Job cards or production runs not invoiced|
|Cash taken by owner||Taxpayer admission|
|Till tapes do not reconcile to bankings|
|Missing sales invoices|
|Loose cash control|
|Low recorded drawings|
|No adjustment for cash taken|
|Additional taxable activity||Interactivity transactions|
|Unexplained source of funds|
|Inadvertent mixing of records|
|Barter transactions||Low volume of sales to stock on hand|
|Assets or stock required without apparent funding|
|Understated debtors||Inadequate debtor records|
|Inadequate listing of debtors|
|Level of debtors inconsistent with size of business or apparent credit activity|
|False claims for input tax credit|
|False invoices||Alterations to invoice copy|
|Photocopied invoices in records|
|Lack of records or data on vendor|
|Capital item purchased cannot be inspected|
|Misdescription of purchase||Inconsistencies in invoice data|
|Evidence of alteration|
|Multiple claims||Copy of invoice used|
|Customs fraud||Mismatch between duty value and credit claimed|
|Fictitious business||Abnormally high refund|
|Hotel or post office quoted as address instead of street address|
|New business claiming large refunds|
The Discriminant Function System is usually called DIF, in tax administration jargon. This is a computer system to select taxpayers for audit. The system should work on mathematical formulas that measure the probability of tax error in taxpayers’ returns. The United States was a pioneer in designing and implementing a DIF for the audit of income taxes.