A. Introduction and Summary

1. A value-added tax (VAT) with a single positive rate of 17 percent was introduced on July 1, 1996 in Uganda after two years of preparation with assistance from the Fund.² It replaced a sales tax on goods and selected services at rates between 12 percent and 30 percent. Several problems began to affect the operations of the VAT soon after its introduction. First, the number of potential VAT taxpayers was much larger than planned; second, resistance to the new tax from traders escalated into a strike in late September to early October 1996. Third, the computer system installed for the tax was not adaptable enough to provide information required for its management. Finally, the response of the authorities to the strike and the priorities of VAT administrators after the strike exacerbated compliance under the tax during the first year of its operation. As a result, compliance was poor during the first few months following the introduction of the tax and remained below expectations during the first year of its operation. This note describes some of the main problems that have caused the poor compliance rate.

2. Although the VAT in Uganda has a single positive rate and broad coverage, its initial threshold of U Sh 20 million may have been set too low, and a number of items that should have been exempted were zero rated. The threshold was increased to U Sh 50 million in November 1996, and the number of zero-rated items was reduced in July 1997. However, the administration of the tax is still complicated by the zero rating of items in the agricultural sector, the treatment of aid-funded projects, and the exemption of fuel. Significant improvements in administrative procedures will be required to reach an adequate level of efficiency. Existing procedures have inadequate features that complicate taxpayers’ compliance. These include a complicated filing and payment system and an ineffective treatment of VAT credits (the offset mechanism). These are burdensome procedures that penalize honest taxpayers; meanwhile, noncomplying payers have been able to avoid their tax obligations without penalty. Furthermore, the computer system did not prove to be operationally effective. The development and implementation of the computer system were also complicated by changes (policy and procedural) that were introduced in attempts to improve performance of the VAT.

3. The introduction of the VAT provides an opportunity to modernize tax administration by developing new systems that are based on modern self-assessment procedures. These systems include using modern taxpayer numbering systems, standardizing invoices and books of accounts, simplifying filing and payment obligations, and implementing effective collection-monitoring systems and comprehensive enforcement programs. Because of administrative capacity constraints, this approach was not taken at the introduction of the VAT. Instead, the administration that was established operates independently of the existing income tax and customs administrations. For example, the income tax and customs files are not fully utilized. In addition, while a single taxpayer identification number has been created for all the revenue departments, a new VAT number was introduced, thereby complicating the exchange of information between concerned departments. In recent months, however, there has been a growing consensus among URA senior officials to share information among the customs, income tax, and VAT administrations, and to implement a more integrated approach to tax and customs administration, including establishment of a large-taxpayer unit.

B. VAT Registration and Compliance

4. Initial studies have not made clear what the size of the VAT register should be. In March 1994, a rough technical study suggested that with a threshold of U Sh 20 million the register would likely contain no more than 3,000 potential payers.³ However, subsequent studies indicated that with such a threshold the register would comprise between 7,000 and 8,500 payers—the size that was expected in July 1996 when the tax was introduced. Despite a comprehensive VAT education program, the registration effort faced some difficulties in getting responses from potential payers, and toward the end of the registration phase many potential filers who had not responded to requests to file were arbitrarily put on the register. The result was that the register on July 1, 1996 numbered 13,500 payers, or 5,000—6,500 more than expected. In order to encourage a reduction of the register to its expected size and an increase in compliance, a filing target of 80 percent of all those registered was set for end-December 1996. However, the traders’ strike in September–October 1996, the subsequent raising of the threshold, the temporary restrictions imposed on the activities of VAT officials by the authorities following public complaints about the VAT, and the VAT administrators’ emphasis (for a couple of months) on collecting VAT from payers who had been deregistered (by virtue of the higher threshold) all worked to reduce compliance during the second half of 1996. Although compliance improved during the first half of 1997, major uncertainties remained about the accurate size of the register even by the end of the fiscal year 1996/97 (July–June). The URA then decided to retain on the register all VAT payers with a turnover of below U Sh 50 million, provided that they complied with filing obligations; at the same time, it “deactivated” some nonfilers, including those above the threshold. As a result, the size of the register at end-September 1997 was 3,448, including all 400 of the largest payers and 1,480 payers below the threshold (Table 1). Even though the size of the register remains somewhat uncertain, the authorities are making efforts to ensure that all taxpayers with a turnover of over U Sh 50 million are on the register, by cross-checking with income tax, customs, and VAT files, by checking lists of those contracting with government and public agencies, and by cross-checking with professional and trade associations. Recent field surveys of taxpayers have also brought onto the register a significant number of VAT payers with a turnover above the threshold.

Table 1.

Uganda: VAT Taxpayers—Distribution by Number and Annual Turnover, September 1997

Source: Uganda Revenue Authority.

Table 1.

Uganda: VAT Taxpayers—Distribution by Number and Annual Turnover, September 1997

		Number of Taxpayers	Percent of Total	Amount of Turnover (in millions of Uganda shillings)	Percent of Total
Below U Sh 50 million		1,480	43	31,377	1
U Sh 50–U Sh 100 million		670	20	46,922	2
U Sh 100–U Sh 250 million		567	16	92,163	4
U Sh 250–U Sh 500 million		275	8	100,671	5
Over U Sh 500 million		456	13	1,944,470	88
	Total	3,448	100	2,215,603	100

Source: Uganda Revenue Authority.

View Table

Table 1.

Uganda: VAT Taxpayers—Distribution by Number and Annual Turnover, September 1997

		Number of Taxpayers	Percent of Total	Amount of Turnover (in millions of Uganda shillings)	Percent of Total
Below U Sh 50 million		1,480	43	31,377	1
U Sh 50–U Sh 100 million		670	20	46,922	2
U Sh 100–U Sh 250 million		567	16	92,163	4
U Sh 250–U Sh 500 million		275	8	100,671	5
Over U Sh 500 million		456	13	1,944,470	88
	Total	3,448	100	2,215,603	100

Source: Uganda Revenue Authority.

C. Filing and Payment Procedures

5. One factor that has inhibited compliance is the onerous nature of the current filing procedures—for the taxpayers, as well as for tax administration. Ideally, taxpayers should be responsible for completing and filing their VAT returns and paying their taxes without the assistance or intervention of a VAT inspector. Under the current procedures, however, the taxpayer must be present at the tax office and at the bank at least once each month. The administration must open and record at least three separate events each month in each account: receipt of the VAT return, issuance of the payment notice, and receipt of the payment at the bank after clearance of the check. Posting of liabilities and payments is also cumbersome, as it involves opening the same account several times each month; moreover, the system is incapable of handling more than one payment per return. What is needed are procedures by which the vast majority of monthly returns and remittances can be completed in a single transaction by the accounting unit, without reference to a tax inspector. Such a system would minimize the resources utilized in tax accounting and free the inspectors to concentrate their attention on those taxpayers who fail to file their returns or pay the appropriate tax on time.

6. Late payment, underpayment, and nonpayment of the VAT has been a problem since the introduction of the tax. The current level of VAT arrears represents approximately 16 percent of the gross domestic VAT collected in the past year (VAT arrears for the largest 400 firms are approximately U Sh 5.9 billion, more than 10 percent of the VAT paid by them in the past year). The VAT law provides for the administrative application of a penalty tax for both late filing of a VAT return and late payment of the tax. To date, only the penalty tax for late filing has been applied. Approximately 80 percent of the taxpayers have filed returns since July 1997.

7. All registered VAT taxpayers must file a VAT return and pay their VAT liabilities by the fifteenth of each month. VAT enforcement officers immediately follow up on all VAT returns outstanding on the sixteenth of each month and begin encouraging taxpayers to pay the tax by the end of the month. If there is no response, this first contact is followed in a few days by a warning from a tax administrator of the penalties if the taxpayer does not comply within the required time. The vigorous collection efforts of the enforcement officers are, however, being undermined by weak penalty and interest provisions—an issue that the authorities are addressing. There is no requirement in the VAT law for taxpayers to pay interest on outstanding taxes. The law provides for a penalty tax to be applied to taxes outstanding at a rate 5 percentage points higher than the official bank rate of the Bank of Uganda. This same rate is payable by the URA for refunds after the due date. At present, this penalty tax rate is approximately 20 percent per annum; the rate at which individuals may borrow money from banks ranges from 18 percent to 24 percent per annum.

8. VAT payers with more tax credits than output tax liabilities in a tax period may request a refund of the excess credit. In such a case, the taxpayer must provide certified copies of all invoices for purchases, import documents, export documents, and a summary of all sales during the claim period. The amount of VAT refunded each month has varied from a high of U Sh 2.29 billion to a low of U Sh 229 million. At present, there is a backlog of approximately 175 unprocessed claims, representing U Sh 1.85 billion. All refund claims are audited prior to payment. Alternatively, taxpayers with excess credits may request to offset (carry forward) the credits against future VAT liabilities. If this option is chosen, no supporting documentation is required, and approval is granted after a desk review, subject to audit at a later date.

D. Audits

9. Audit activities were limited to the auditing of refunds until June 1997 (Table 2). Two divisions carry out audits: the Refund Audit Section and the Ordinary Audit Section. The first section employs 14 staff and deals with special refund claims (particularly those of large taxpayers), which represent approximately 20 percent of total VAT refunds. The number of claims audited by this section increased recently—from approximately 30–35 claims a week in October 1997, to 70–80 in November 1997. The Ordinary Audit Section comprises 25 officers. Its target is to carry out 75 audits a month, 50 percent of which would be on the largest taxpayers. The VAT and Internal Revenue Departments have only recently attempted to perform some joint audits.

Table 2.

Uganda: Ordinary VAT Audits, July–December 1997

(In millions of Uganda shillings)

Source: Uganda Revenue Authority

Table 2.

Uganda: Ordinary VAT Audits, July–December 1997

(In millions of Uganda shillings)

				Outcome
	Number of Ordinary VAT Audits			(In millions of Uganda shillings)
Month	Top 400	Other	Total	Assessment	Refunds
July	10	36	46	536.3	1.5
August	20	47	67	860.0	158.9
September	15	43	58	1,436.9	124.9
October	22	74	96	1,287.8	282.3
November	20	43	63	1,100.9	466.4
December	12	23	35	1,381.5	5.0
Total	99	266	365	6,603.4	1,039.0
Net assessments				5,564.4

Source: Uganda Revenue Authority

View Table

Table 2.

Uganda: Ordinary VAT Audits, July–December 1997

(In millions of Uganda shillings)

				Outcome
	Number of Ordinary VAT Audits			(In millions of Uganda shillings)
Month	Top 400	Other	Total	Assessment	Refunds
July	10	36	46	536.3	1.5
August	20	47	67	860.0	158.9
September	15	43	58	1,436.9	124.9
October	22	74	96	1,287.8	282.3
November	20	43	63	1,100.9	466.4
December	12	23	35	1,381.5	5.0
Total	99	266	365	6,603.4	1,039.0
Net assessments				5,564.4

Source: Uganda Revenue Authority