15 Refunds

Liam Ebrill, Michael Keen, and Victoria Perry
Published Date:
November 2001
  • ShareShare
Show Summary Details

The logic of the VAT requires that some taxpayers, particularly exporters, receive refunds. The practical problems this creates—often among the most troublesome that arise in implementing the VAT—are the topic of this chapter.


It is a key feature of the invoice-credit form of VAT that some businesses will pay more tax on their inputs than is due on their output, and so ought in principle to receive refunds. This will be true, in particular, of exporters (since their output is zero-rated, so long as the destination principle applies), and those whose investment purchases are large relative to their current sales (a group that may well include new enterprises). These are especially important groups in terms of wider economic development. Without effective treatment of excess credits, the VAT runs the risk of introducing significant and costly distortions for these groups.140 At the same time, however, the payment of refunds evidently can create lucrative opportunities for fraud and corruption. Moreover, financially strapped governments may be tempted to delay refunds for sheer need of cash. As discussed in the previous chapter, an effective refund system is also closely connected to the successful implementation of a comprehensive VAT audit program.

In principle, refunds should be paid promptly following the first VAT return after the excess credit arises, and should be paid irrespective of the nature of the enterprise or the circumstances giving rise to the credit. That is indeed typically the case in most developed countries: refunds are usually paid within about three to four weeks of the end of the taxable period. Many developing and transition economies, however, limit the entitlement to refunds. Hard information on this is sparse. Table 15.1 reports on practice in 27 survey countries and FAD advice on somewhat fewer. It should be noted, however, that the practice reported in this table is typically that laid down in law or regulation, which experience suggests can differ quite markedly from that actually implemented.

Table 15.1.The Treatment of Excess Credits
FAD Recommendation (number of respondents)Practice (number of respondents)
Refunds to exporters only (with or without lag)78
Refunds to exporters immediately; all others with lag66
Carry forward (for all) for some period321
Refunds to exporters and those making large capital investments only01
Year-end (or quarterly) adjustment only11
Refund in excess of a floor credit size only14
Not given in practice, although provided for in the law01
Completely ad hoc01
Source: IMF staff compilation.

Of which one refunded only after audit.

Source: IMF staff compilation.

Of which one refunded only after audit.

Clearly many countries focus their refund efforts on exporters. The means by which this is done vary. Nearly one-quarter of respondents give refunds to no other group. FAD has also advised in some cases (for example, Gabon) that imported foreign capital equipment be exempted for certain investors.

There is some evidence that FAD advice on refunds has become less “purist” in recent years: increasing awareness of the risks of fraud, particularly in transition countries, has led to a wider acceptance of departures from the theoretically ideal practice of immediate refunding. The results of the survey show that, in most developing countries, FAD now typically recommends limiting refunds of excess credit to exporters; other taxpayers are thus requested to carry forward their excess VAT credits during at least six months. While this approach is appropriate to deal with most taxpayers whose sales are made predominantly on the domestic market, there is no doubt that in a few cases this recommendation underestimates the difficulties faced by some enterprises with significant excess credits. This may be the case for major investments (with sometimes the business activity beginning several months after the investment is made) or in case of structural credit position due to a dual rate structure (outputs being taxed at the reduced rate while most inputs are taxed at the standard rate).

A critical aspect of the treatment of refunds is the overall VAT audit strategy, including the systems adopted to verify refund claims. Hard evidence is limited to relatively few countries. However, as discussed in the previous chapter, the general impression is that the lack of appropriate audit programs in many countries contributes to the ineffectiveness of the refund system, which relies heavily on extensive pre-refund checks prior to granting refunds rather than the selective post-refund strategy typically recommended by FAD.

A poor refund system can do significant harm. Failure fully to refund excess credits undermines the core principle of the VAT: intermediate transactions between firms should bear no final VAT burden. Such failure creates distortions that imply a real waste of resources: producers will substitute away from taxed inputs, and relative prices between sectors will be affected. The competitiveness of the export sector may be harmed, and the competitive edge tilted against new firms.

The Magnitude of the Issue

The refunds required under an invoice-credit VAT can be substantial, although the appropriate level of refunds will vary across countries according to their circumstances (see Box 15.1). In 1994–95, for instance, the United Kingdom collected over £68 billion in VAT payments and returned over £26 billion in refunds—that is, over 60 percent of net collections. While this is unusually high, reflecting the extensive zero-rating in the United Kingdom, refunds are extensive in many other countries too. In France, for example, VAT refunds amount to 40 percent of the net collections, while in Sweden they are nearly 80 percent; in Korea refunds run at around 50 percent of net VAT receipts.

Reasons for Failure and Possible Responses


While the appropriate treatment of excess credits in theory is clear—prompt refund—a number of obstacles to this arise in practice. Even when all parties are honest and capable, some transaction costs unavoidably imply some imperfection in the process. Three more considerations also arise.

Consider first the potential for fraud. The point is nicely put by Bird (1993), noting that a VAT is “… uniquely susceptible to fraud to the extent that a supplier’s invoice (or export certificate) in effect constitutes a check drawn on government [and so] inherently constitutes a tempting target for those who would loot the treasury.” By concealing sales to the domestic market, traders not only evade the requirement to charge tax on their output but, moreover, enable themselves to reclaim tax on their input. Overstatement of exports, sometimes with the support of falsified customs documents, can be a particularly attractive way to do this, in that it apparently explains the use that has been made of inputs. Of course, when refunds are restricted to exports, as is the case in a number of countries, it is precisely the assertion that items have been exported that is needed to open the possibility of refund. Temporary shell corporations, false export documentation with respect to actual production sold domestically, falsification of invoices, production and purchase records, and misreporting of prices are all used extensively to support claims of excess credits to be refunded. In the EU, for example, various types of fraud have developed under the VAT regime, with the more complex devices emerging with the recent elimination of the borders between EU members for the circulation of goods, persons, and funds. Other types of fraud have also developed in transition economies in central Europe. In this connection, the capacity of countries such as Poland and Hungary to develop effective VAT audit programs and control the movement of goods on their Eastern borders will clearly be a major challenge for their accession to the EU.

Box 15.1.What Is a Reasonable Level of Refunds?

It is important for tax administrations to have a sense of the level of refunds they might reasonably expect to pay, both as a guide to the level of resources that the activity might require and as a warning signal of attempted fraud.

Broadly speaking, refunds as a proportion of (net) VAT revenues under a fully functioning, single-rate, invoice-credit VAT would be

where i and z denote respectively the shares of investment and zero-rated items (including exports) in GDP, α is the proportion of investment that generates excess credits (a quantity one would expect to be higher in faster-growing economies), λ the ratio of value added to sales in the zero-rated sector and e the efficiency ratio. Refunds will clearly be higher (a) in faster-growing economies (both because investment will be higher and more of it is likely to be by new firms with no output tax to take credit against), and (b) in more open economies.

The refunds required can quite plausibly be very large. Suppose, for example, that one is considering an economy in which the investment ratio is 10 percent, and exports account for 40 percent of GDP. Assuming that 5 percent of investment generates excess credits and that value added is 40 percent of sales in the export sector, achieving an efficiency ratio of 35 percent—around the average for a BRO country—will imply paying refunds equal to about 70 percent of net collections.

Second, the power to make refunds provides opportunities for corrupt practices by tax officials. Payment might be extracted in return for prompt provision (or overprovision) of refunds. Anecdotal evidence in some developing countries suggests that there are countries in which there is a widely known price for obtaining refunds of, say, 5 percent of the amount refunded. Customs officials might also connive in issuing false export documentation.

Short of outright corruption, moreover, the need for refunds creates scope for favoritism in the treatment of particular firms or sectors. There have been cases, for instance, in which complaints are voiced that domestic enterprises receive refunds more readily than foreign.

Third, particularly in transition countries with no history of voluntary tax compliance as well as in the many other countries that find themselves short of revenue, governments may find delaying the payment of refunds a tempting source of funds. By delaying payment, the government may be able to extract a forced interest-free loan from taxpayers.141 In some cases, such practices have resulted in a major reaction of the exporter community. In recent years, complicated exemption and zero-rating schemes applied to exporters and their suppliers (so called indirect exporters), have been implemented in a few developing countries (mainly in Asia), in an attempt to deal with cash flow problems resulting from excessively slow refund of excess VAT credits. Similar schemes also apply in some developed countries’ VAT systems (such as the Irish and the Netherlands VAT systems; see below). However, while EU countries can afford the administrative burden of such complex mechanisms, developing countries with weak administrative capacity clearly cannot. Indeed, in those developing countries, such schemes send the wrong message to the business community regarding governments’ commitment to simplifying and enforcing the tax system, with the immediate result of further degrading the design of the VAT and significantly complicating its operations.


Some of these difficulties evidently extend far beyond the context of VAT. Dealing with corruption, in particular, is a matter for wider civil service and other reforms. Within that wider setting, however, there are a number of approaches that might be taken to achieve an acceptably smooth system of refunds.

Offsetting excess VAT credits against VAT and other tax arrears

Some experts would argue that excess VAT credits (refunds) should never be offset against taxes other than future VAT liabilities, as is sometimes done. Such cross-offsetting would result in loss of control over data on taxpayers in all the affected taxes. This view may be excessively rigid for most developing and transition economies, where tax administrations have to cope with significant noncompliance problems, including, in many cases, substantial tax arrears. In these countries, the most common practice is to ensure that a taxpayer has no outstanding tax arrears (from VAT and other taxes) before processing VAT refunds.

Limiting entitlement to refunds

Scarce administrative capacity might be focused on the most important refund claims by simply denying some kinds of refund that, in principle, ought to be allowed. The simplest such rule is a de minimis requirement specifying an absolute amount, which must be exceeded before a refund is granted. This, however, will inevitably bear proportionately more heavily on small (perhaps new) firms rather than large.

A preferable approach would be to require excess credits to be carried forward for some specified time (three to six months, for example) and pay a refund only if they remain unrecovered. This is rough justice: a credit delayed is to some degree a credit denied. But it may be an acceptable means of excluding less important excess credit positions from the requirements for refund monitoring. In some contexts, tax administration may be so weak that even this exposes the revenue to too much risk. One might then require the indefinite carry forward of all excess VAT credits against future VAT liabilities.

Carry forward, however, is unlikely to prove satisfactory for exporters—who will find themselves permanently carrying forward unrelieved excess credits—or for firms that incur considerable costs prior to beginning operations. Even a requirement to carry forward for some fixed period effectively implies a charge equal to the interest foregone in deferring credit for the period of the carry forward, which can be a substantial sum.

Treatment of exporters

Many countries make special arrangements for exporters. These arrangements seek to ensure that exporters of good repute, at least, obtain prompt refund. This might involve a “gold card” scheme, for instance, under which exporters with good payment records obtain repayment within some specified period; silver card holders, and those in any lower groups, receive somewhat less prompt treatment. However, these schemes create their own difficulties and distortions. They create administrative complications in requiring some apportionment of credits for enterprises that both export and sell domestically. Moreover, schemes that offer beneficial treatment of those with established reputations inevitably disfavor firms that may be honest but are simply too new to have built up such a record; this, however, seems to be an inevitable consequence of the optimal audit strategy.

The refund processing system for exporters recommended by FAD is based on a few simple principles,142 namely:

  • distinguishing between refund claimants with a history of compliance and those claimants with poor or unknown compliance histories;

  • using pre-refund audits for high risk refund claims and postrefund audits for claims of lesser risk;

  • the application of criteria to determine the likely extent of revenue risk associated with each refund claim;

  • maintenance of historical profiles for each refund claimant; and

  • the performance of refund audits that focus only on verifying the facts of the refund claimed.

VAT excess credits related to import of capital goods

While recommendations regarding the prompt treatment of refunds in developing countries are usually targeted to exporters, efforts have also been made recently to better deal with excess credits due to large capital expenditures. VAT exemptions for investors who import large amounts of capital equipment have been limited to very few countries. Indeed, this practice may be easily abused and result in pressures to extend VAT exemptions to other imports. As described in Box 15.2, the development of a simple deferred-payment scheme for large amounts of imports of capital goods has recently been seen to be a more appropriate approach.

Audit strategy

A well-designed and executed audit strategy is critical to the construction of a smooth refund system. Many countries rely heavily on extensive pre-refund auditing, which is clearly a misuse of audit resources. In such cases, the vast majority of VAT audit resources are absorbed in the verification of VAT refunds as a result of the requirement that every claim be audited prior to approval. This preoccupation with the pre-refund audit of all claims leaves very few audit resources available to implement a comprehensive audit program. Even strong tax administrations would typically be unable to provide timely clearance of all refunds in this way.

Box 15.2.Guidelines for a Deferral System for VAT on Imported Capital Goods

A deferral system for the VAT on imports should have the following characteristics:

  • The scheme should be limited to importers of large amounts of capital goods who are registered VAT taxpayers.

  • Capital goods, both imported and domestic, are subject to the standard rate VAT.

  • Imports of capital goods by persons who are not registered VAT taxpayers are subject to VAT at the time of import. VAT is paid, as usual, before the clearance of goods.

  • Importers of capital goods who are registered VAT taxpayers are permitted to defer accounting for the VAT until their next VAT return.

  • In this return, the value of the capital goods and the VAT applicable to those goods are reported as a VAT liability, and, in the same return, the appropriate VAT input tax credit is taken for the capital goods.

  • If the importer is entitled to 100 percent input tax credit (equipment used exclusively in taxable activities) the VAT applicable to the importation, reported as a liability, will be completely offset by the corresponding input credit.

  • The customs office is furnished a copy of the VAT return to close their records of the importation.

The strategy recommended by FAD envisages prerefund audits only for first and unusually large refund claims, with emphasis concentrated instead on selective post-refund audits. Claims for refunds in excess of a certain level should be subject to quick, issue-oriented refund audits. New Zealand, for example, assigns each taxpayer a limit below which refunds will be made automatically. However, insofar as they are known to taxpayers, such regularities in audit may facilitate some evasion devices; even dishonest taxpayers, for instance, will ensure that their first refund claim is accurate. “Gold card” schemes of the kind described above also provide an incentive to acquire a reputation for honest dealing, and then abuse it. This is clearly an area where all risks of fraud cannot be entirely eliminated—the success of any audit and refund programs depends on the capacity of the tax administration to adapt its methods and procedures to taxpayers’ changes of behavior (that is, they depend, ultimately, on the good judgment of experienced, skilled auditors).

To limit the risks of abuse, any refund system that limits pre-refund audits to a few cases—those with the highest risks for tax revenue—needs to be complemented by an effective audit program. This program needs to have a broad coverage and to include taxpayers for whom the refunds are usually processed without prerefund verifications. Information collected during these audits, especially data on refunds that were processed without pre-refund audits, are critical to updating taxpayers’ profiles and the criteria used for the processing of refunds.

Removing input tax on those likely to have excess credits

An alternative strategy is to remove the need for refunds altogether by eliminating tax on the inputs of enterprises most likely to run up excess credits. As mentioned above:

  • Certain countries effectively zero-rate supplies to exporters. In Ireland, for instance, those who export more than 75 percent of their output can obtain an authorization that enables their suppliers not to charge VAT. In Korea, those who supply exporters (referred to as “indirect exporters”) are zero-rated in respect of selected transactions if they can produce an authorized letter of credit relating to the final export.

  • A number of transition economies and some developing countries exempt from VAT the import of foreign capital equipment, at least by foreign investors. This precludes ensuring that the VAT is collected at the easiest point, import, but, in principle, achieves the same substantive result (assuming output tax is charged appropriately by the investor on later sales) as imposing the tax at the border, and crediting it and refunding excess credits in a timely manner.143

These devices, however, do break the VAT chain, and so significantly increase the complexity of the VAT, and, eventually, expose revenue to major risks.

  • If supplies to exporters are zero-rated, it becomes necessary to ensure that final output is not diverted to the domestic market tax-free. Moreover, if indirect exporters consequently find themselves in an excess credit position, the problem is merely shifted back one stage perhaps to even more enterprises, and ones typically harder to control. In Korea, extending the logic of effectively zero-rating indirect exporters has led to similar zero-rating, in some circumstances, for the second or third preexport stage, complicating control problems still further: indeed it has been suggested that these provisions underlie the high ratio of refunds to net receipts in Korea.

  • If exemption on capital purchases is to be granted only to foreign investors, it becomes necessary to guard against fraudulent shell corporations set up offshore by domestic persons to take advantage of the exemption. Breaking the chain in this way can also distort economic activity: exemption for capital goods imported by established foreign producers, for instance, distorts the system by lowering costs in those instances, relative to the use of domestically manufactured capital equipment, and relative to domestic owners. Extension of such import exemptions to eliminate the competitive distortion, or exemptions for all capital equipment purchases, simply extends the opportunities for fraud and abuse.

When tax administration is weak, breaking the chain in these ways is likely to be unduly risky. And when it is not weak, breaking the chain is unnecessary.


The refund system is one of the principal pressure points in the administration of the invoice-credit VAT. This is obviously an area in which tax policy advice has been strongly influenced by tax administration constraints over the past years, recognizing that implementation of best practices, including immediate refund of all VAT excess credits, was simply not possible in countries with weak administrative capacities.

In many developing and transition countries, the lack of effective audit programs is certainly a main factor explaining the difficulties faced in implementing sound collection procedures and effective refund systems, based on self-assessment principles. During the early part of the renewed spread of the VAT over the past ten years, it seems that technical assistance recommendations have been rather successful in dealing with the risks of weak collection procedures, especially when effective debt collection systems have been implemented for large taxpayers in an LTU. However, during the same period, it may be that the capacity of tax administrations to develop effective audit programs and, hence, to provide timely refunds without inviting excessive abuses has been overestimated.

In recent years, however, audit issues and refund problems have become a major concern in technical assistance work and this can be expected to continue. FAD’s current strategy to improve the operation of refunds is to limit entitlement to refunds, by ensuring prompt refunds to exporters (and, more recently, to enterprises importing large amounts of capital goods), while imposing some delays on others. This, however, does compromise the basic nature of the VAT, which requires that all excess credits be refunded without loss of value in order to avoid competitive distortion.

As discussed in Chapter 14, future improvements in current refund practices are subject to the development of appropriate audit strategies: better audit selection and methods, better coordination of income tax, VAT, and customs operations, increased training of auditors, and stronger tax administration management and supervision. However, this is an area in which achieving decisive progress is difficult, and refunds are likely to be a continuing concern in some developing countries and transition economies for some time. The difficulty of dealing with refunds has proved, to some degree, the Achilles’ heel of the VAT.

    Other Resources Citing This Publication