Book
Share
Chapter

6 A Survey of Advice and Experience

Author(s):
Liam Ebrill, Michael Keen, and Victoria Perry
Published Date:
November 2001
Share
  • ShareShare
Show Summary Details

Many important issues arise in designing and implementing a value added tax. While the headline rate of the tax naturally attracts most attention in public discussions, issues concerning the base and administration can be no less important in determining the impact of the tax. Detail on such matters, however, is relatively hard to come by; not all will even be contained in the law, and there may in any event be a gap between rules and practice. To develop a picture of the pattern of VATs implemented in developing and transition countries, staff in the IMF’s Fiscal Affairs Department (FAD) were asked to complete detailed questionnaires on a number of countries in which the department has provided technical assistance in relation to the VAT over the past decade or so. Questions covered not only actual practice, but also the advice that FAD had provided on these items: the intention here is to identify any areas in which “best practice” is systematically not adopted, since this would raise the possibility that the thinking behind these standard recommendations ignores considerations that loom large in practice. Of interest too are cases in which advice varies, suggesting either a sensitivity of design issues to country circumstances or an underlying intellectual issue that remains unresolved.

The Survey

The questionnaire sought detailed information on the broad context of the VAT (such as its yield and the kind of tax it replaced), FAD recommendations, and current practice as regards a range of policy and administrative matters. Broadly usable responses were obtained for the 37 countries listed in Table 6.1, although item nonresponse was quite high for several questions, especially on details of tax administration. Most questionnaires were completed in spring/summer 1998.

Table 6.1.The Countries Surveyed
Start of VATSub-Saharan AfricaAsia and PacificCentral Europe and BRONorth Africa and Middle EastAmericas
Before 1980Bolivia (1973)

Peru (1973)
1981–90Philippines (1988)

Pakistan (1990)
Haiti (1982)
1991–95Benin (1991)

Burkina Faso (1993)

Gabon (1995)

Mauritania (1995)

Togo (1995)

Zambia (1995)
Bangladesh (1991)

China (1994)
Azerbaijan (1992)

Georgia (1992)

Krygyz (1992)

Moldova (1992)

Russia (1992)

Poland (1993)

Bulgaria (1994)
Jamaica (1991)

El Salvador (1992)

Venezuela (1993)
1996–99Guinea (1996)

Uganda (1996)

Mauritius (1998)

Tanzania (1998)

Cameroon (1999)

Ghana (1999)
Sri Lanka (1997)

Mongolia (1998)

Cambodia (1999)

Vietnam (1999)
Albania (1996)

Croatia (1998)
PlannedFYR Macedonia (2000)Yemen
Note: Date in parentheses refers to the year of introduction.
Note: Date in parentheses refers to the year of introduction.

The sample was not, and was not intended to be, random. Rather it principally reflected the geographical scope of FAD’s activities in relation to the VAT over the previous ten years, and capitalized on the detailed knowledge of available staff. The result was a particular emphasis on African countries (about one-third of the sample) and transition countries (also around one-third). Reflecting the importance of pre-implementation involvement, the sample included six countries which, at the time of the survey, did not yet have a VAT in place. Since the sample reflected advice given at various points over the previous ten years (though the “practice” described in the survey responses is intended to be current practice), recorded variations in advice given, or implementation strategy adopted, could reflect changes over time in perceived “best practice” rather than systematic differences in opinion or the tailoring of advice by circumstances.

Consistent Advice Consistently Followed

The survey showed that FAD’s advice in the past decade had been quite consistent across countries for many of the tax policy and administration issues explored in the survey. Further, in many of these areas of consistency, actual practice broadly conformed to FAD advice. These seemingly uncontentious prescriptions include that: (i) the VAT should utilize the invoice-credit method; (ii) the VAT should extend through the retail stage; (iii) services should be included in the base of the tax except for well-specified exemptions for education, health, and financial services; (iv) the introduction of a VAT is facilitated by the establishment of large taxpayer units (LTUs);63 (v) the VAT should be administered as a national tax; (vi) in addition to incorporated businesses, individual entrepreneurs and unincorporated businesses should all be covered by the VAT, subject to a threshold to exclude the smallest businesses; and (vii) a preparation and implementation timetable and project team devoted solely to VAT implementation should be established before a new VAT is introduced.

Consistent Advice Not Consistently Followed

Areas in which FAD advice was consistent but was not consistently adopted include:

  • Zero-rating: for all survey countries for which there is information (32),64 FAD recommended that exports be zero-rated, with trade being taxed on a destination basis (that is, where the good is consumed). In practice, exports were zero-rated in 28 countries; the exceptions occurred in the CIS, which adopted a mixed origin/destination basis depending on whether trade was among themselves or with the rest of the world. However, 12 out of 22 surveyed countries responding on this point also zero-rate items other than exports.

  • Input credits for capital goods: in 28 of 30 cases, it was recommended that VAT charged on capital goods be credited in full immediately. In eight surveyed countries, however, full and immediate credit was not given for capital inputs.

  • Number of rates: in all but two cases, the advice was for a single positive VAT rate. In practice, almost one-third of surveyed countries (7/22) set more than one rate.

  • Cash versus accrual basis: for all 28 surveyed countries, the advice was that the VAT be calculated on an accrual basis (meaning that tax is payable, broadly speaking, when effective delivery or transfer of ownership occurs). In practice, an accrual basis is used in 27 of 32 surveyed countries with 5 on a cash basis.

  • Staffing and training: VAT staffing was seen as appropriate in only 73 percent of the surveyed countries (11/15). More stark, however, is that training was reported as appropriate in only 25 percent of surveyed countries (6/24).

Variations in Advice

On several issues, advice varied noticeably across countries:

  • Threshold: the FAD-recommended threshold below which businesses need not file for the VAT varies quite widely. More striking and troubling is that the survey reveals a strong tendency for countries to adopt thresholds significantly lower than those recommended, with an associated higher number of registrants.

  • Exemptions and their control: in line with international practice, FAD recommends that certain services, principally health, education, and non-fee financial services, be exempt from the VAT for policy and practical reasons. In 20 countries, FAD reports have recommended permitting some additional exemptions of various types. And in virtually all cases, nonstandard exemptions exist in practice, particularly, though not exclusively, including imports and supplies related to diplomats, NGOs, donor-funded projects, and incentives provided under investment codes. FAD generally advises against the broader exemptions that have emerged in practice.

  • Rate of tax: the standard rate of VAT recommended by FAD has varied widely, clustering between 11 and 19 percent. This variation is not surprising given the differences in needed revenue and levels of economic activity.

  • Agriculture: FAD recommended exempting agricultural production in 13 cases, and taxing it in 10 cases. Practice is, however, consistent—of the 28 countries for which there is survey information, 25 exempt agricultural production, while only 3 transition countries subject it to tax; 13 of the 25 countries also exempted inputs to agriculture.

  • Organization of the VAT administration: FAD typically recommends that the domestic component of the VAT be administered by the internal tax department (that is, administered with the income tax in a function-based organization65). This is currently the case in 80 percent of the surveyed countries (29 out of 36 countries). However, in a few countries that administer indirect taxes in the customs department (mainly former British colonies), FAD advice was that the VAT be administered by the existing customs department or that a new VAT department be established.

Impediments to Administrative Implementation

The survey results indicated that there were a number of key implementation areas where FAD’s advice or administration of VAT had been adopted, notably, in setting the VAT implementation strategy, ensuring taxpayer education, and collection enforcement (including through large taxpayer units). However, in some areas, countries had set out to follow FAD advice on administrative aspects of the VAT but failed to implement them satisfactorily. Reported problem areas included:

  • weak procedures to register taxpayers in some countries;

  • absence of self-assessment (whereby taxpayers calculate and pay their tax liability directly) in some countries. Moreover, even in countries that do have self-assessment, filing and payment procedures are often excessively cumbersome;

  • poorly functioning refund systems—delays in processing exporters’ refunds claims are particularly worrisome; and

  • insufficient audit coverage and inadequate audit methods.

Conclusions

There are clearly several essentially noncontentious issues in VAT design. More striking is that there are several areas in which FAD advice has been consistent but not widely accepted. These include the wider-than-recommended use of zero-rating; the lack of the recommended timely input credits for capital goods; and the frequency with which multiple-rate VATs are adopted. Perhaps the most notable, however, is the tendency for countries to adopt thresholds lower than recommended. There are also areas in which there has been variability in both advice given and accepted—notably, as regards exemptions and the treatment of agriculture. Finally, the survey results identify a number of tax administration impediments (see Table 6.2) to the successful implementation of the VAT. These points are taken up in the chapters that follow.

Table 6.2.Implementation of IMF Tax Administration Advice: Summary
Number of ResponsesAdvice Fully Implemented (appropriate)Advice Partly Implemented (medium)Advice Partly Implemented (insufficient)Advice Not Implemented
VAT preparations
VAT implementation team30291
Implementation timetable261961
Involvement of customs299128
Staffing15114
Training246117
Registration275184
Education241095
VAT administration
Use of self-assessment procedures31265
Forms (and procedures) easy to complete (and to comply with)2713104
Accuracy of taxpayers’ accounts2411112
Detection of nonfilers268117
Penalties3110174
Collection of penalties24699
Functioning of VAT refund system19478
Audit coverage21219
Audit methods213810
Large taxpayer units31247
Source: IMF staff tabulation.
Source: IMF staff tabulation.

    Other Resources Citing This Publication