Maldives: Technical Assistance Report––Estimating Tax Expenditures

The Maldives has identified the estimation and regular reporting of tax expenditures (TEs) as one of the top priority areas in continuing its tax modernization process. TEs are alternative policy tools (e.g., to direct transfers and other spending measures) in the form of provisions in the tax legislation that modify the tax liability of individuals or companies. The cost of TEs should be identified, measured, and publicly reported to improve transparency in fiscal management.

Abstract

The Maldives has identified the estimation and regular reporting of tax expenditures (TEs) as one of the top priority areas in continuing its tax modernization process. TEs are alternative policy tools (e.g., to direct transfers and other spending measures) in the form of provisions in the tax legislation that modify the tax liability of individuals or companies. The cost of TEs should be identified, measured, and publicly reported to improve transparency in fiscal management.

Executive Summary

The Maldives has identified the estimation and regular reporting of tax expenditures (TEs) as one of the top priority areas in continuing its tax modernization process. TEs are alternative policy tools (e.g., to direct transfers and other spending measures) in the form of provisions in the tax legislation that modify the tax liability of individuals or companies. The cost of TEs should be identified, measured, and publicly reported to improve transparency in fiscal management.

The report focuses on TEs in the Goods and Services Tax (GST) and Business Profit Tax (BPT) in the Maldives. TEs in the recently adopted Personal Income Tax (PIT)—that entered into force in 2020—are left for future capacity development (CD) work as the necessary data are still in the process of collection for the first time in the Maldives.

In particular, the report summarizes the results of assisting the Maldives in:

  • i) identifying the BPT and GST Benchmark tax systems: A key issue for the estimation of TEs is to determine the reference tax system (known as the “Benchmark” system), thereby viewing any deviation from it as a TE. For the GST, the Benchmark system is a uniform GST rate (except for a zero-rate on exports) with no GST exemptions. For the BPT, generally, deviations from the BPT benchmark system include any reduced tax rates, tax deductions for expenses that are unnecessary for doing business (such as deductions for charitable donations and deductions for employee welfare expenses), and BPT exemptions. Moreover, to enhance transparency and guide policy making, the mission recommends costing and publishing the revenue implications of several additional measures that are part of the Benchmark as memo items, such as limitation to interest deductions and the higher tax rate on banks.

  • ii) building TEs models: The mission delivered (i) a BPT model based on business-level tax return data in 2019; and (ii) a GST model based on the Supply and Use Tables (SUT) from the national accounts (using the latest available SUT data from 2014). Both TE models build on static microsimulation techniques and hence (beyond computing TEs) can help in simulating the direct revenue effects of hypothetical tax policy changes. Concrete examples are provided in the report—in the context of (i) arbitrage opportunities between the PIT and BPT; and (ii) and tax incentives.

The mission delivered a hands-on training on using both TE models to a joint team from the Ministry of Finance (MoF) and the Maldives Inland Revenue Authority (MIRA). The mission recommends designing the institutional setting for estimating TEs in the Maldives to ensure continuity and integrity of the TE estimates. For example, individual ministries and governmental bodies should not be able to manipulate the definitions of TEs and/or the data and methodology used for estimating their costs. It is essential to form a devoted team (initially a joint MoF-MIRA task force headed by the MoF) to assume the responsibility of estimating TEs and updating the models; establish the needed process for obtaining data regularly (from the National Bureau of Statistics, MIRA, and potentially other data providers); and establish effective knowledge management.

Developing capacity for estimating TEs is a continuous process. The main purpose of this stream of CD work is to provide technical support and training enabling the authorities to further improve the delivered models, rather than presenting the estimates of TEs per se. The devoted TE team should advance the delivered models and gradually expand the estimation of TEs to other areas such as the PIT and analysis of the international aspects of corporate taxation.

Preliminary estimates of TEs in the Maldives are set out in the report (and a high-level summary is presented below). These estimates, however, are subject to important caveats as laid out in detail in the report. For example, TEs in the GST require finetuning various assumptions and updating the data as new SUTs become available. Another important caveat is the uncertainty surrounding the impact of COVID-19 on revenues in 2020. The largest TE item in the BPT is the exemption threshold. The largest TE item in the GST is the deviation of the non-tourism rate from the tourism rate (6 percent versus 12 percent in 2020, and 6 versus 8.66 percent in 2014). Total TE takes into account interaction between individual items of TEs (i.e., it is not the sum of TEs from individual items).

It is important to emphasize that the exercise of estimating TEs is neither a normative analysis nor a compliance gap analysis. The effectiveness and efficiency of TEs in achieving their intended goals require a separate devoted analysis from the mechanical estimation of TEs. Similarly, measuring revenue not collected due to noncompliance requires a revenue administration gap analysis, and is beyond the exercise of estimating TEs. Furthermore, TEs are not comparable across countries due to different assumptions about the Benchmark and differences in the designs of the tax system.

Summary of Recommendations

article image

Summary of Estimates of TEs

article image
Source: IMF staff calculation.Note: This table provides the results of costing of TE and memo items for the BPT and GST in the Maldives. Values are shown for the most recent year with actual data (2019 for BPT and 2014 for GST) and a projection for 2020. Numbers are reported in MVR millions and as a percentage of the corresponding year’s total tax revenue. The table reports the revenue impact (cost) of the deviation of a provision from the Benchmark (TE item) and memo items. A positive number for a TE item indicates an increase in revenue from removing the deviation of this item from the existing system. Memo items, here, show the additional revenue raised in the system due to the specific provision. For 2014, the benchmark GST rate is 8.66 because it was raised in November from 8 to 12 percent. In 2020, the benchmark GST rate is 12 percent.

I. Introduction

1. Revenue forgone from tax expenditures (TEs) should be measured and regularly published to improve transparency in fiscal management. TEs are alternative policy tools (e.g., to direct transfers and other spending measures) in the form of provisions in the tax legislation that modify the tax liability of individuals or companies. In line with international best practice, many countries annually (or bi-annually) estimate and publicly report the revenue forgone from TEs (OECD, 2010).

2. The Maldives plans to regularly publish TEs in major taxes. In a welcome first step, a pilot study has reported TEs that arise in import duty. However, TEs in the Goods and Services Tax (GST) and Business Profit Tax (BPT) have not yet been estimated and published.

3. This report summarizes the results of assisting the Maldivian authorities in building TE models for the BPT and the GST. The models were delivered with a hands-on training to a joint team from the MoF and the Maldives Inland Revenue Authority (MIRA). Beyond presenting the underlying methodology, the purpose of the training was to enable the attendees to further develop the models and point to potential areas for future improvement. The material provided during the workshop describes in detail the functioning and operation of the models. Therefore, the report presents a high-level summary. Since the first Personal Income Tax (PIT) in the country just entered into effect in 2020, the necessary data for estimating its TEs are not yet fully available.

4. The report proceeds as follows. Section II describes the BPT Benchmark and the model for estimating TEs in the BPT. Section III describes the GST Benchmark and the model for estimating TEs in the GST. Finally, Section IV briefly discusses the reporting of TEs.

II. Tax Expenditures in the Business Profit Tax

A. Benchmark of the Business Profit Tax

Overview

5. Since 2019, the Maldives has achieved remarkable progress in improving tax policy by adopting several measures.1 These include: (i) establishing a Tax Policy Unit (TPU); (ii) introducing a PIT; (iii) abolishing a preferential tax regime (in the form of a reduced rate of 5 percent on foreign income); (iv) abolishing all tax exemptions and incentives under the foreign investment act (FIA); (v) introducing transfer pricing rules and limitation to interest deductions; and (vi) introducing a cross-border withholding tax (WHT) on all sources of capital income as well as technical and management fees.

6. As of 2020, the Maldives taxes business income of firms at a rate of 15 percent through the Business Profit Tax Act (BPTA). The standard BPT rate of 15 percent applies to a relatively broad based, with a few exceptions. A tax-free threshold of MVR 500,000 applies to all taxpayers. Financial accounting profit is generally recognized with some adjustments, such as a fixed 20 percent deduction for rental income, deductions for charitable donations, and deductions for welfare expenses on employees. There is an earning stripping rule that limits interest deductions to 30 percent of EBITDA. Prior to 2020, a minority of taxpayers were eligible for a reduced rate of five percent (0.5 percent of BPT filers).

7. Banks are taxed at 25 percent and transitioned from the Bank Profit Tax to the Business Profit Tax in 2020, while maintaining the higher 25 percent tax rate. Prior to 2020, the Bank Profit Tax was legislated through the Bank Profit Tax Act and accompanying regulations. The computation of taxable income slightly deviated from that stated in the BPT.

8. The PIT on individual’s remuneration was implemented in 2020 through the Income Tax Act (ITA). The PIT applies to earnings, including remuneration, capital income and dividends, received by individuals and partnerships resident or established in the Maldives. The tax follows a progressive rate structure with zero percent tax applying to earnings up to MVR 720,000 and a top rate of 15 percent that applies to income above MVR 2.4 million.2 By choosing the legal form, businesses that previously paid the BPT (e.g., as partnerships) can elect to pay the PIT as sole proprietorships.3 Roughly 26 percent of BPT payers in 2019 will pay PIT in 2020.

Benchmark

9. The benchmark for BPT is a single tax rate and it excludes tax reliefs other than those for necessary business expenses. Moreover, aspects of the tax system that are designed to improve its functioning, such as anti-tax avoidance rules, are included in the Benchmark.4

10. The mission identified a handful of provisions that deviate from the Benchmark in the 2019 and 2020 BPT regimes. The Benchmark BPT regime Identified TEs in 2019 and 2020 are:

  • I. A reduced tax rate of 5 percent tax for specific taxpayers (this item was removed in 2020);5

  • II. A tax-free threshold of MVR 500,000;

  • III. Other deductions that are not related to the cost of doing business, which are: o Deductions for charitable giving (Section 31 of the BPT Act);

    • o Employee welfare expenses (Section 30 of the BPT Act); and

    • o Excess GST used to offset BPT payments (although MIRA and MOF staff stated that de facto no offsets occurred in 2019 or 2020, but in principle the law allows for this offset).

11. In addition, Zakat-al-Mal could also be considered a TE. Zakat-al-Mal is a mandatory payment of 2.5 percent of the stock of net wealth (if a minimum level of wealth, called Nisaab, currently set at MVR 6,961.50, is held for the entire lunar year). Zakat-al-Mal is deducted for the calculation of the BPT, if paid to government. Since it is mandatory, it could be argued that it is part of the Benchmark (e.g., some countries such as Canada allow for deducting the property tax against the corporate income and this is regarded as the Benchmark). However, it is not directly linked to the business expenses and the deduction reduces the BPT liability. Moreover, in practice individuals may opt not to pay Zakat-al-Mal to the government.6 Thus, the deduction for Zakat-al-Mal has the feature of a TE, and is regarded in this report as a TE. Whether as a TE or a memo item, the mission recommends computing and publishing the budgetary implication of deducting Zakat-al-Mal payment for the BPT.

12. Arguments can be made for and against classifying the higher tax rate on banks (of 25 percent) as a deviation from the Benchmark. It can be argued that the higher rate is part of the Benchmark because banks fell under a separate law (and filed a separate tax return) until 2020. Starting from 2020, banks are taxed under the BPTA with a rate of 25 percent. Some countries with a similar higher rate do consider it as a Benchmark. For example, the United Kingdom imposes an eight percent additional tax rate on banks and includes this higher rate in the Benchmark. However, it can also be argued that the increased tax rate for banks is a negative TE since it deviates from what should be a uniform rate of 15 percent. While typical TEs are positive since they reduce BPT collection, the higher rate on banks, if considered as a TE, increases BPT collection (and hence leads to a “negative” TE). Regardless of whether the higher rate for banks is deemed as the Benchmark, it is recommended to report the additional revenue raised from the higher tax rate on banks, whether as a TE or memo item.

13. Beyond TEs, the mission recommends estimating the revenue implications of several memo items and publishing them in the TE report. These include the: (i) fixed rental income deduction; (ii) limits on head office expense deductions; (iii) limits on non-monetary remuneration deductions; and (iv) earnings stripping rule. Additionally, the higher tax rate on banks should be reported as a memo item (as done in this report) if it is not classified as TEs.

B. Business Profit Tax Data

14. The first step in the estimation of TEs is obtaining and preparing the data. MIRA maintains a fairly good quality BPT database at the taxpayer level that contains all detailed information from the tax returns, including the economic sector of the business. With minimal cleaning and converting (steps were described and documented in the workshop), the data can be entered as inputs in the TE model. The mission used primely anonymized taxpayer data from 2019. Until 2019, banks filed tax returns separately from the BPT, and thus needed to be added to the database.7 Starting from 2020, all companies file under the ITA.

15. Analyzing the distribution of taxpayers across firm sizes and industries is critical to understand TE results and guide policymaking. In total, income tax payments contributed 19 percent to total tax revenue in 2019 (3.5 percent of GDP). This section provides detailed summary statistics using business-level anonymized BPT tax return data.

16. In 2019, there were over 10,000 taxpayers under the BPT and Bank Profit Tax across 23 industries, with tax payments concentrated in a few large taxpayers. Entities with turnover exceeding MVR 100 million paid 75 percent of total BPT but only made up 2.4 percent of filers (Figure 1). However, there was a significant share of small filers with positive tax paid (1.2 thousand). About 70 percent of those filers with positive taxable income had turnover of less than MVR 20 million. The majority of filers had turnover of less than MVR 500,000 (i.e., below the tax-free threshold), and many reported nil turnover (49 percent).

Figure 1.
Figure 1.

Distribution of Business Profit Taxpayers, by Turnover

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation.

17. Similarly, BPT payments were concentrated in a few industries. The sectors with the largest BPT payments were tourism (29 percent), banks that paid Bank Profit Tax (21 percent), and trade and vehicle repair sectors (14 percent) (Figure 2). Thus, the tourism and banking sectors paid about 50 percent of the total business tax revenue, although they comprise a relatively small number of firms (7 percent of total number of firms). At the same time, about 35 percent of filers are in the trade and vehicle repair sector. All other sectors paid 37 percent of total BPT combined and made up 58 percent of BPT filers. In 2019, PIT filers paid four percent and non-bank current BPT filers paid about 75 percent of the income tax (Figure 3).

Figure 2.
Figure 2.

Distribution of Business Profit Taxpayers, by Industry

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation.
Figure 3.
Figure 3.

Non-Banks, Banks, and Current Personal Income Tax Files

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation.

C. Model for Estimating Tax Expenditures in the Business Profit Tax

General Structure

18. The mission team built a static model tailored to the Maldives to calculate TEs in the BPT. In principle, TEs are computed by using available data, say in year t (2019 in this case), and applying the tax system of the same year (historical TEs) or by applying the tax system of a future year (say in t+1, t+2,…). The latter computation is a projection of TEs. The following general steps describe the estimation of TEs in the model:

  • I. Estimate the 2019 revenues from 2019 BPT returns database. Despite that actual 2019 revenues are known, the model estimates the revenue in 2019 to ensure that the revenue difference between the existing tax system and a modified one (e.g., the Benchmark) is only attributed to modifications to the tax code. The model estimation of 2019 BPT revenue under the existing system almost matches actual revenues (with the fit being at 99.9 percent).

  • II. Estimate tax paid under the Benchmark BPT system for each identified TE item individually.

  • III. Define the TE for a particular item as the difference in tax paid between both tax systems (i.e., the revenue forgone method). This exercise (steps I-III) has been done for 2019 (i.e., applying the 2019 systems to the 2019 data) and 2020 (i.e., applying the 2020 systems to 2019 data, with adjustments to approximate changes in the tax base since 2019 as described in the accompanying workshop).

  • IV. Steps I-III can be done for each category of TE separately (in which case the sum of individual TEs does not consider any interaction between them) or sequentially taking into accounts the interaction between TE provisions.

19. Figure 4 visualizes the structure of (and the various steps in) the Excel-based model using the Excel sheet names where each calculation occurs. The model calculates tax payable for each firm under a given BPT tax regime and then aggregates all the firm-level BPT payable to compute total BPT payable. The steps to do so is as follows: firm-level data on revenue and each deduction from the 2019 BPT tax returns is saved in the “database”.8 Then, a given tax regime, with parameters coming from the “dashboard”, is applied to firm-level data to calculate taxable income and BPT payable in the “tax calculator”. The results are aggregated across all firms and shown at the sector- and economy-level in the “dashboard” and “summary statistics”. Results for several tax regimes, including the Benchmark regime, are compared in the “dashboard” sheet.

Figure 4.
Figure 4.

Business Profit Tax Model Structure

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff.

20. The BPT model is flexible allowing for the adjustment of a variety of parameters to evaluate changes to the BPT tax regime. These parameters cover individual TE items (discussed below) and other key elements of the BPT tax regime. The user can analyze the revenue impact of a tax regime with any combination of values for each parameter. A comprehensive list of the parameters is provided in Table 1.

Table 1.

Parameters in the Business Profit Tax Model

article image
article image
Source: IMF staff.

21. It is important to note that the TE methodology captures only the direct revenue impact of the deviation from a Benchmark or changes to the BPT tax regime (i.e., a “static” model).9 This in line with common practice of estimating TEs. The model does not account for behavioral responses from taxpayers and government or other economic impacts caused by changes to the tax regime. While results should be interpreted and communicated with this caveat, the methodology still provides a powerful tool to assess the direct revenue impact of TEs. It can be complemented with other economic models and qualitative information to account for dynamics and general equilibrium effects.

22. The mission provided a hands-on training on the use of the BPT model. In particular, the workshop explained in detail all the steps taken to clean the data and place them in the database sheet; the structure and steps within the tax calculator sheet; and how the model emulates the tax return.

First Results — BPT Expenditure for 2019 and 2020

Results for 2019

23. The mission calculated TE for both 2019 and 2020 by costing TEs in the existing BPT regime that deviate from the Benchmark tax regime discussed above. TEs are calculated and reported for each individual TE item to increase transparency and guide decision-making. This calculation is done by comparing the tax collected from the existing tax regime to that of the same existing regime after removing a single TE item (e.g., no tax-free threshold while retaining all other TEs in the existing system).

24. The sum of individual TEs will not equal the total TEs because it ignores interactions between the individual TEs. For example, the elimination of the tax-free threshold increases tax liability more if all taxpayers are assumed to pay 15 percent tax, but the TE from the tax-free threshold is calculated assuming that the existing tax rates are unchanged (i.e., some taxpayers pay a 5 percent rate). It is important to clearly communicate the impact of interactions when reporting TE to avoid confusion and to account for interactions when costing multiple changes to the BPT.

25. Results suggest that the 2019 BPT TE is MVR 284 million (9 percent of total BPT collected or 1.7 percent of total tax revenue) if the higher tax rate on banks is not counted as a TE item (Figure 5). The largest item is the tax-free threshold, while the 5 percent tax rate and charitable giving deductions are relatively small as few taxpayers with positive taxable income pay the 5 percent rate or make use of deductions for charitable giving (MVR 49 million in charitable giving deductions were claimed in 2019). The TE from Zakat al-mal deduction is also low—below 1 percent of BPT revenues.

Figure 5.
Figure 5.

Business Profit Tax Expenditure in 2019

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation.

26. If we consider the higher tax rate on banks as negative TE, the 2019 BPT TE is MVR 27 million (one percent of total BPT collected or 0.2 percent of total tax revenue) (see Appendix). The 25 percent tax rate on banks raises significant additional revenue (40 percent more than would be collected under the standard rate of 15 percent).

27. There is significant variation in the allocation of TE across industries (Figure 6). The trade and repair of vehicles sector benefits most from TEs due to a large number of relatively small taxpayers, all of which benefit from the tax-free threshold that disproportionately supports small businesses. The tourism sector has fewer, but larger, taxpayers and makes significant deductions for employee welfare expenses. Conversely, if the higher 25 percent tax rate on banks was classified as a TE, banks would have a large negative TE.

Figure 6.
Figure 6.

Business Profit Tax Expenditure by Industry in 2019

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation.

28. Revenue forgone (or additional revenues) from memo items should be also computed and reported to enhance transparency and guide decision making. Figure 7 presents the budgetary implications of memo items in 2019, all of which led to higher revenues. These include the: (i) earning-stripping rule; (ii) limits on deductions for non-monetary remuneration; (iii) limits on head office expenses; (iv) fixed rental income deduction; and (v) 25 percent rate for banks paying the Bank Profit Tax. Note that in Figure 7, a positive amount is interpreted as a reduction in BPT revenue, while a negative amount is the revenue raised from that provision. The revenue impact of the analyzed items other than the limit on non-monetary remuneration deduction and higher tax rate on banks are small. Banks were excluded from the analysis for most items due to the unavailability of information.

Figure 7.
Figure 7.

Costing of Business Profit Tax Memo Items

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation.

Results for 2020

29. While projections generally entail a degree of uncertainty, estimating TEs for 2020 is particularly challenging due to the significant impact of COVID-19. The 2020 BPT TE was calculated by applying the 2020 BPT regime—excluding taxpayers that no longer pay BPT under the ITA—and adjusting for changes to the tax base. The necessary caveats should be clearly communicated when reporting 2020 TEs. Moreover, estimates should be updated once actual 2020 BPT data become available.

30. To project 2020 BPT payments, modifications to the 2019 tax base were done by reducing variable revenue and cost items for sectors that are expected to be more impacted by COVID-19 (all except financial services and information and communications). The adjusted variable revenue and cost items were revenue from the financial statement, rental income, cost of goods sold, and expenses incurred to generate rental income. The level of the adjustment (64 percent of 2019 levels) was calibrated so that overall BPT revenue fell by 46 percent, which is in line with the IMF World Economic Outlook (WEO) projections for income tax collection in the Maldives for 2020.

31. Results suggest that the 2020 BPT total TE was MVR 60 million (4 percent of total BPT collected or 0.7 percent of total tax collection)if we ignore the higher rate on banks (Figure 8). The BPT TE reduction from 2019 to 2020 (9 to 4 percent of total BPT collected) can be explained by the tax-free threshold having a reduced impact on BPT tax collection since many small taxpayers pay the PIT instead of the BPT under the ITA. Relatedly, the trade and repair of vehicles sector has a reduced TE benefit due to the lower impact of the tax-free threshold (Figure 9). If we consider the higher tax rate on banks as negative TE, the 2020 BPT TE was estimated to be negative MVR 197 million (negative 12.8 percent of total BPT collected or negative 2.4 percent of total tax collection).

Figure 8.
Figure 8.

Business Profit Tax Expenditure in 2020

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation.
Figure 9.
Figure 9.

Business Profit Tax Expenditure by Industry in 2020

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation.

Microsimulation—Illustrative Examples of Costing Hypothetical Policy Changes

32. Beyond estimating TEs, the model can also be used to assess the direct revenue impact of possible policy changes to provisions in the BPT tax code. The static methodology is based on comparing the revenue collected under the existing system to that collected under an alternative regime. Recall, the model is designed to evaluate the impact of changing any of the parameters in Table 1, and can in principle be extended to account for more policy scenarios.

33. To illustrate how the model could be used for static microsimulation purposes, we consider two exercises. The first one estimates the revenue implication of businesses that choose to be taxed under the new PIT in 2020 instead of the BPT. The second exercise estimates the revenue costs of hypothetical tax incentives.

Migrating from the BPT to the PIT

34. Disparity in the taxation of incomes under different legal forms generates arbitrage opportunities. In this exercise, we estimate the revenue implication from differences between the PIT rate scale and that of the BPT— this differential leads to lower tax payments under the PIT for some businesses. According to the data, in 2020, about 26 percent of businesses migrated from the BPT to the PIT. The tax payment under the PIT can be simulated based on available information from the BPT return.

35. The net revenue impact of the migration of some businesses from the BPT to the PIT is MVR 44 million (1.4 percent of total BPT revenue). This is effect is decomposed as: (i) a negative revenue income under the BPT of MVR 128 million; and (ii) additional revenue of MVR 84 million under the PIT.

Tax Incentives

36. This exercise estimates TEs for specific examples of hypothetical tax policies to incentivize investment. The costed policies are (i) tax holidays provided to the top 1 or 10 percent of taxpayers based on reported taxable income; (ii) a one-time 15 percent non-refundable tax credit for investment (aircrafts, marine vessels, earth moving vehicles, and plant and equipment); (iii) a one-time 15 percent non-refundable tax credit for Research and Development (R&D)—as captured by intangibles and computer software; and (iv) and a one-time full expensing of all capital spending. Banks were excluded from the analysis due to a lack of information.

37. This TE exercise is not a cost-benefit analysis of tax incentives. Evidence suggests that tax holidays usually are ill-designed and tend to be an inefficient and ineffective policy tool (for a full discussion, see IMF, 2015)—especially in light of increasing popularity of minimum taxes in the home countries of parent companies that invest abroad and strengthening controlled foreign corporation (CFC) rules. Both, minimum taxes on outbound investment and CFC rules generally trigger taxes in the resident country if income in the source country is not taxed at a minimum level. This implies that a tax holiday generates revenue in other countries and becomes less effective in lowering the tax of multinational enterprises (MNE). In contrast, empirical evidence suggests that well-deigned temporary cost-based tax incentives in the form investment tax credits or full expensing are effective in accelerating investment (see, e.g., Wen, 2020).

38. The capital spending for firms in 2019 is estimated using the following methodology. For firms with tax return data for 2018 and 2019 (roughly 80 percent of firms in the 2019 BPT return), capital spending is calculated as the total cost of assets in 2019 (inclusive of assets disposed of during 2019) minus the total cost of assets in 2018 (exclusive of assets disposed of in 2018). If this resulted in negative asset purchases for a given firm and asset category in 2019, then asset purchases for that category and firm is set to zero. For firms without tax return data for 2018, capital spending is calculated as the total cost of assets in 2019 less accumulated depreciation on those assets (this is an approximation and assumes that all assets―less cumulative depreciation―were purchased in 2019).

39. Figure 10 presents the estimated direct revenue costs of these tax measures. Tax holidays offered to one percent of large taxpayers (100 taxpayers in total) would significantly reduce BPT collection and are extremely costly. This is because a few large taxpayers pay the vast majority of BPT. In practice, the eligibly to a tax holiday may imply a different number of taxpayers but would still present significant revenue risks due to the high share of tax paid by few large taxpayers.

Figure 10.
Figure 10.

Microsimulation—Costing of Hypothetical Tax Provisions

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation.

40. The revenue loss from full expensing and tax credit is large, but lower than that from a tax holiday. Taxpaying firms in the Maldives have little investment in intangibles and computer software, leading to a lower revenue impact from a policy that provides credits for such costs. Meanwhile, costs for physical assets are large (estimated to be MVR 7 billion in 2019, ten times larger than that of intangibles and computer software). Full expensing would provide immediate tax relief for profitable firms making capital investments.

41. One important caveat is that this analysis was done using a single period model, which understates the revenue loss from tax credits and overstates the loss from full expensing. Tax credits provided to companies making losses are generally carried forward to subsequent years or refunded, leading to reduced revenue in future years (estimated to roughly double the fiscal cost of the tax credit options when not accounting for the time value of money). While full expensing decreases revenue in the first (and potentially immediately succeeding years),10 it would increase revenue in medium-term as firms could no longer claim depreciation deductions on the fully expensed items. A full accounting of tax credit and expensing policies would require a multiple period model with net present value analysis.

III. Tax Expenditures in the Goods and Services Tax

A. Benchmark of the Goods and Services Tax

Overview

42. At present, the GST in the Maldives makes a distinction between the tourism sector (with a GST rate of 12 percent) and non-tourism sectors (with a rate of 6 percent). With its present basic design features, the GST in the Maldives is in some key aspects (but also with an important exception) comparable to the many multistage invoice-credit based Value-Added Tax (VAT) systems adopted globally. The most important deviation from a standard VAT is the separation between the tourism sector and all other sectors of the economy. Section 14 of the GST Act states that GST tax chargeable under the law is categorized into two separate categories: (a) tourism goods and services; and (b) general goods and services apart from those under (a). The Act proceeds to define which goods and services are to be considered tourism and which are not, and the tax rates to be applied to each of these categories. The GST Regulations (Chapter 1, section 3(c)) specifies that if a person carries on taxable activities in both areas, these must be reported to MIRA separately. However, in practice, sales between tourist and non-tourist companies appear to be treated under normal VAT rules concerning charging and crediting VAT.

43. The average (revenue weighted) GST rate is about 8.8 percent. About 63 percent of GST revenues is from the tourism sector, reflecting the importance of tourism for the Maldivian economy and the difference in the GST rates. In 2019, GST revenue was 8.7 percent of GDP.

Benchmark

44. The Benchmark of the GST is a uniform GST rate of 12 percent applied to all goods and services consumed in the Maldives, with no exemptions (except for public goods), and with no zero-rated goods or services (except on exports). Thus, the Benchmark assumes that the GST is effectively imposed on final consumption. Zero-rated exports are part of the Benchmark because the VAT design should be neutral with respect to its effect on relative prices and international trade (by treating all exports, as well all imports, alike for tax purposes consistent with the destination principle). The VAT Benchmark requires that at each stage of production, businesses can claim tax credits to recover the GST paid on their business inputs or capital investments, thereby applying the GST only to the value added at each production stage (ensuring neutrality with respect to investment).

45. GST paid by the government and its entities on their inputs is considered part of the Benchmark. Also, goods and services produced by the government, such as education or health services, are exempted from GST and these exemptions are also considered part of the Benchmark. Governments and non-profit organizations cannot claim input tax credits to recover the GST paid on inputs used to supply goods and services that are not subject to the GST, which is regarded as part of the Benchmark according to international best practice.11

46. Hence, deviations from the GST Benchmark in the Maldives are:

  • I. The application of a rate of 6 percent on goods and services outside the tourism sector (vs. the GST rate of 12 percent rate).

  • II. Exempt goods and services in the GST Act (listed in the Appendix).

  • III. Zero-rated goods and services in the GST Act (listed in the Appendix).

47. There are three specific aspects in the Maldivian GST that, in principle, comprise a deviation from the Benchmark but have not been explicitly modelled in the TE model that is delivered in conjunction with this report:

  • I. The GST registration threshold: Suppliers below the turnover level of MVR 1,000,000 (roughly corresponding to USD 64,500) fall below the GST registration threshold, thereby generating (i) TEs (as GST is not paid on final demand); and (ii) negative TEs (as GST paid on inputs of these suppliers is not recovered). However, in practice, in the Maldives, the net TE below the GST registration threshold is expected to be very low. First, the data show that registered taxpayers with turnover close to the threshold contribute very little to the GST (Figure 11). Second, importers of goods to the Maldives and suppliers of tourism goods and services are required to register, even if the value of their supplies does not exceed the limit of MVR 1 million.12 Third, in line with international best practice, the present legal provisions allow for businesses with turnover under the registration threshold to voluntarily request to register with the MIRA for GST purposes. This allows a number of smaller businesses to register for GST in cases when this is advantageous to them (e.g., if they primarily make supplies to larger registered businesses).

  • II. No refund where input tax exceeds output tax: The input tax can be only set off against the output tax and there are no refunds if it exceeds the output tax. Although the Law allows for offsetting excess GST against the BPT, in practice, it seems (according to the authorities) that such offsetting is rather rare. To the extent that taxpayers are claiming input tax credits in the future, it becomes a timing issue. However, this does imply a negative TE on any excess of input tax over the output tax that is not refunded or reclaimed by the taxpayer, but this is not captured in the TE estimates.

  • III. Delays in claiming input tax credits on capital expenditures (CAPEX): Claiming the input tax on CAPEX exceeding 500,000 MVR, in the form of setting off against output tax, is only possible gradually over 36 months rather than immediately. This timing issue ultimately generates a negative TE in a particular year that is not estimated in the model due to the lack of necessary data.

Figure 11.
Figure 11.

Distribution of Goods and Services Tax Payments Based on Turnover Thresholds

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation based on data from MIRA (provided in 2019). Numbers on the horizontal axes are in million MVR.

B. Goods and Services Tax Data

48. The Supply and Use Tables (SUTs) are the main source of data to estimate GST revenues under the existing and the Benchmark systems. The TE model uses the latest available SUTs for Maldives, which cover 2014,13 and provide disaggregated information for 40 industrial sectors and 51 commodities. The SUTs are used to estimate: (i) the level of GST paid by final consumers (i.e., final demand by households, governments, and non-profits organisations); (ii) exports consumed by tourists in the Maldives; and (iii) the amount of GST paid by businesses on the purchase of intermediate inputs and capital investments used to produce goods or services that are exempted from GST. These purchases do not receive refund on the GST paid.

49. The estimation of GST revenues also requires the calculation of effective tax rates (ETRs) associated with each commodity purchased by domestic consumers or by tourists.14

ETRs are the product of two components “tax rate x taxable share”, as follows:

  • I. The tax rate is 12 percent on purchases of goods and services by tourists; zero for the zero-rated or exempted goods and services; and 6 percent for all other goods and services purchased by domestic consumers.

  • II. For each commodity, the taxable share, and non-taxable (due to an exemption or a zero-rate) should be obtained or computed. These taxable shares are different for household’s consumption and for exports (consumption in the Maldives by tourists).15 For example, if, for a given commodity, 50 percent of the expenditures is zero-rated and 50 percent is subject to a 6 percent tax rate, the overall ETR for this commodity is (50% × 6%) + (50% × 0%) = 3 percent.

C. Model for Estimating Tax Expenditures in the Goods and Services Tax

General Structure

50. The workshop provided training and material explaining in detail how to use the GST model for estimating TEs. Therefore, only a compact high-level summary is provided below.

51. The GST model for estimating TEs has the following general structure (see also Figure 12):

  • I. Data input:

    • o SUTs.

    • o The share of each commodity that is exempted or zero-rated for the GST.

    • o The share of each commodity that is allocated to exports.

    • o ETRs applicable to each commodity.16

  • II. Estimate GST on final demand from the Use Matrix (as summarized in the formulas in Box 1), for:

    • o Household and NPISH expenditures.

    • o Governments expenditures.

    • o Exports (for goods and services consumed in the Maldives).

    • o Ideally, the modeler should disentangle government/public capital formation, residential construction, and business/private capital formation. GST is paid on government and residential construction, but for business capital formation, GST paid will be refunded unless the business produces exempt goods or services.

  • III. Estimate non-refundable GST paid on business inputs in sectors producing exempt goods and services. Businesses or other organizations that produce GST exempt services are not eligible to claim input tax credits on purchases made towards the provision of these services. Therefore, purchases of such inputs form part of the GST tax base. For each industrial sector, the unrecoverable GST is equal to the multiplication of business input expenditures from each commodity with its ETRs and the share of exempt supplies calculated for this industrial sector.

    • o From the Supply Table: Compute the share that is exempted from the GST by multiplying the share of exempt commodity by the production of these exempt commodities. For example, 85 percent of the commodity “financial services” is deemed to be exempted. Thus, all sectors producing the commodity financial services (e.g., banks or insurance companies) are characterized with a large portion of the GST paid on their inputs that is not subjected to an input tax.

    • o From the Use Table: Compute the amount of GST paid on business inputs. The sum of the GST paid for a given industry is then simply multiplied by the share of exempt supply.

    • o Ideally, the modeler should disentangle the amounts of gross fixed capital formation (GFCF) by industrial sector to account for unrefunded GST on business GFCF. However, due to data unavailability, the model assumes that the GFCF is allocated to each industrial sector based on their share of consumption of fixed capital, which is one of the elements of the value-added presented in the Use Table. For example, if the Electric Power Generation sector represents 6.7 percent of the total consumption of fixed capital, 6.7 percent of the construction commodity presented in the final demand of the Use Table under GFCF will be allocated to this industry.

  • IV. Having estimated GST revenues under existing system (following the above steps II and III), note that the estimation entails a margin of error. The model currently overestimates total GST revenues by about 3 percent in 2014.

  • V. Redo the estimation for the benchmark tax system.

  • VI. The total GST TE is the difference between GST revenue in the existing system and the benchmark system.

  • VII. For years subsequent to the year of the SUTs, use projections and growth rates to estimate GST revenues and hence TEs. Data sources include GST revenues and macroeconomic projections from national sources and international sources such as the IMF World Economic Outlook (WEO).

Figure 12.
Figure 12.

The Structure of the Goods and Services Tax Model

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff.

Results

52. The total TE in the GST is estimated considering all TE items and is not equal to the sum of individual TE items. TEs can be estimated together (assuming the removal of all deviations from the benchmark at the same time and taking into account interactions between TEs) or can be estimated independently, i.e., removing one TE and comparing the estimated GST revenues under this alternative scenario with the estimated revenues under the existing system. To illustrate an interaction effect, consider for example the removal of the GST exemption for electricity. Under the existing system, the electricity would be taxed at 6 percent. However, under a system with a benchmark tax rate of 12 percent, the estimated revenue forgone from this exemption will be much larger.

53. Figure 13 presents the results from estimating TEs in 2014 and 2020, respectively. The largest individual TE item (whether in 2014 or 2020) is the deviation of the non-tourism GST rate form the Benchmark rate of 12 percent. The second largest item of TE is the zero-rating. The Appendix lists results by categories of goods and services, and by category of taxpayer. Available SUTs are fairly old (for 2014) and thus may not capture changes in the structure of the SUTs and projection of the SUT table entails uncertainty. However, this source of uncertainty regarding the 2020 results is unlikely to affect the results that the largest TE is due to the reduced rate. The increase in the TE in 2020 (compared to 2014) is in part explained by the growth of the non-tourism sector and the higher difference in the rates (12 versus 6 percent in contrast to 8.66 versus 12 percent as explained in Box 1).

Figure 13.
Figure 13.

Tax Expenditures in the Goods and Services Tax in 2014 and 2020

Citation: IMF Staff Country Reports 2021, 234; 10.5089/9781557755131.002.A001

Source: IMF staff calculation. The benchmark GST rate is 12 percent for 2020 and 8.66 percent for 2014.

Formulas for Estimating Goods and Services Tax Expenditures in the Model

  • For domestic final demand (households, non-profits organizations and governments):

    • o Define

      EffectiveTaxRate=TaxableShare×GSTrate
    • o TE for an item i is:

      TEiDomestic=BenchmarkExisting
    • o Existing system:

      TaxableShareExisting×GSTrateExisting×Expenditure1+EffectiveTaxRate(2014)
    • o Benchmark system:

      TaxableShareBenchmark×GSTrateBenchmark×Expenditure1+EffectiveTaxRate(2014)
  • For exports:

    TEiExport=BenchmarkExisting
    • o Same formulas as above, but with

      • a different GST rate: Specifically, the used GST rate for 2014 is 8.66 because the GST rate was raised in November 2014 from 8 to 12 percent. In the later years, the GST rate has been 12 percent.

      • different taxable shares.

  • From unrefunded GST paid by businesses on intermediate inputs and capital purchases

    • o TE for an item i is:

      TEiBusiness=BenchmarkExisting
    • o Existing system:

      TableShareExisting×GSTrateExisting×Expenditure1+EffectiveTaxRate(2014)×ESSBenchmark
    • o Benchmark system:

      TableShareBenchmark×GSTrateBenchmark×Expenditure1+EffectiveTaxRate(2014)×ESSBenchmark,

where ESS stands for Exempt Supply Share (i.e., exempt supply/total supply) by industry.

54. The priority for upgrading the estimates of TEs in the GST is to use updated SUT data as they become available. Moreover, various assumptions and aspects of the GST Model should be revisited, including:

  • I. Taxable shares:

    • o The taxable share of the various commodities in the SUTs (that are either exempted, zero-rated, taxable at 6 percent, or taxable 12 percent) should be updated to reflect more recent information.

    • o Updating these shares requires close collaboration with the National Bureau of Statistics.

  • II. Shifting of GFCF by industrial sector:

    • o The model should be finetuned to account for the share of residential construction in the construction commodity.

    • o The allocation using depreciation of fixed capital should be revisited.

  • III. GST registration threshold:

    • o The model can be finetuned to account for the registration threshold. For example, the GST taxpayer database can be matched with the business income taxpayer database to analyse the overall value-added (essentially wages and salaries plus profits) of unregistered businesses versus value-added of all businesses by sectors that are subject to the GST. For example, if the matching of the GST and BPT data reveals that value-added of unregistered businesses in the Arts, entertainment and recreation & Other service activities industry (data from the BPT) represents 2 percent of the overall value-added of this sector (again using data from the BPT), then we should probably adjust accordingly the taxable proportion of commodities produced by this sector (i.e., the commodity “other services”).

  • IV. Input tax credit refunds

    • o The GST model assumes no delays in the refunds. If the amount of input tax credit exceeds the amount of GST collected, the model assumes that the business will receive a refund for the overpaid GST. This can result in understating the GST if refunds are not issued and the overpayment cannot be used to offset other taxes owed.

  • V. GST and CAPEX

    • o Input tax credits associated to business CAPEX that exceed MVR 500K cannot be claimed immediately but rather it is claimed over 3 (or more) years. The current model assumes that GST paid on capital expenditures can be claimed immediately as an input tax credit and it can be refunded. The model should be finetuned to reflect the timing issues associated to the claiming of input tax credits on these large CAPEX.

    • o The impact of these delays should be, in principle, considered as negative tax expenditures as under the benchmark tax system, the input tax credits associated with all CAPEX should are claimed immediately and refunded if necessary.

  • VI. Growth factors

    • o The model employs growth rates using aggregated GST data, based on taxpayer databases. These growth projections should be reviewed and updated regularly.

IV. Reporting Tax Expenditures

55. Installing the appropriate institutional setting for reporting on TEs is critical to ensure the integrity of the estimates. The institutional setup should (i) enable the relevant team (typically the tax policy unit at the MoF) to access and validate the necessary data on a regular basis; and (ii) ensure that the estimation and reporting of TEs are shielded from any potential manipulation of the definitions of TEs or the data and methodology used for the estimation. Given that the TPU is at early stage of development, a joint MoF-MIRA task force (headed by the MoF) can assume the responsibility of estimating TEs and updating the models. The analytical capacity of the TPU should be further developed.

56. With a few exceptions, countries report annually on TEs, typically in conjunction with the annual budget. Many countries adopt a legal requirement to produce TEs. In terms of coverage, the Maldives is recommended to start reporting annually the TEs in the BPT and GST and gradually expand to other taxes. Importantly, as data become available, TEs in the PIT should be estimated and reported. Another area for future work is the international tax aspect—e.g., including an estimate of the revenue implication of deviations from the statutory non-resident withholding tax rate (WHT) on cross-border interest, dividends, rents, royalties, and management fees.

57. TE reports should include for each item at least: title and brief description, legal reference, type of tax, and type of measure. This is illustrated in Box 2 that presents an example from a particular item (charitable donation) from Canada TE report (2021). The reporting includes in addition: the source of data, a reference to the estimation model, and the number of beneficiaries (i.e., the number of taxpayers that benefit from this TE item, which could be also expressed as a ratio to total taxpayers in this tax category), inter alia. The report includes historical TE estimates and projections (denoted by (P)).

58. For effective communication of the meaning of the presented TEs, it is recommended to caveat the estimation assumptions in the beginning of the TE report. TEs in many countries, including in the models discussed in this report, are computed using a static method. This means that the estimated direct revenue effect assumes that all other factors remain unchanged—i.e., no account of potential change in taxpayer’s behavior, or general equilibrium effects on the macroeconomy (e.g., a particular tax expenditure in the income tax may affect the level of consumption). Moreover, several countries include background tax statistics in the TE report.

59. The exercise of estimating TEs is neither a normative analysis nor a compliance gap analysis. The effectiveness and efficiency of TEs in achieving their intended goals require a separate devoted analysis from the mechanical estimation of TEs. Similarly, measuring revenue not collected due to noncompliance requires a revenue administration gap analysis, and is beyond the exercise of estimating TEs. Furthermore, TEs are not comparable across countries due to different assumptions about the Benchmark and differences in the designs of the tax system.

Recommendations

  • Further build the capacity of the TPU.

  • Estimate and report TEs on an annual basis.

  • Set up the institutional arrangement to protect the integrity of the estimates of TEs, and

    • o Form a joint MoF-MIRA task force (led by the MoF) mandated with maintaining the needed data and further improving the estimation of TEs.

    • o Ensure access to reliable data regularly.

    • o Develop a template for reporting TEs in line with international best practice.

  • Gradually expand the TE estimation and publication to cover additional areas, including:

    • o the personal income tax.

    • o revenue impacts of international aspects (e.g., from cross-border withholding taxes).

Example from Canada – Charitable Donation in the 2021 Tax Expenditure Report

Deductibility of Charitable Donations

article image

Cost Information:

article image
Source: Canada (2021).

Appendix 1. Details for TEs in the GST

List of Commodities in the SUTs that Have a Share that is Zero-Rated in Households and NPISH Expenditures

article image

List of Commodities in the SUTs that Have a Share that is Exempted in Household and NPISH Expenditures

article image

Telecommunications services were exempted until May 1,2014.

Share of Exports Subject to the TGST

article image

List of Exempt Goods in the GST ACT

  • Electricity;

  • Water;

  • Postal Services;

  • Sewerage;

  • Services complementary to electricity, water, sewerage and postal services;

  • Education

  • Health Care

  • Drugs and Medical Devices

  • Sale of goods received as donations by a non-profit body

  • Financial Services

  • Rent from immovable property

  • International transportation services

  • Day care services

List of Zero-Rated Goods in the GST ACT

Essential goods listed in Schedule 1 of the GST Act:

  • Goods and services exported from the Maldives

  • Transfer of business as a going concern

Essential goods listed in Schedule 1 that are zero-rated includes:

  • Rice, sugar and flour

  • Salt

  • Milk

  • Cooking oil

  • Eggs

  • Tea leaves

  • Deep sea fish, reef fish, all types of fish packed in the Maldives, and rihaakuru

  • Potatoes and onions

  • Ingredients used in making curry paste (cumin, fennel, coriander seed, turmeric, garlic, ginger, chili, chili powder, cinnamon, cardamom, peppercorn, and any other such ingredient)

  • Dhiyaahakuru, kaashi, kurun’baa, rukuraa, and kurolhi

  • Carrots, cabbage, beans and tomatoes

  • Fruits

  • Bread, buns and rusk

  • Baby food

  • Baby diapers and adult diapers: all kinds of baby diapers and adult diapers, including cloth diapers, are zero-rated

  • Cooking gas, diesel and petrol: cooking gas, diesel and petrol are zero-rated. However, other fuel oils such as kerosene and jet fuel are not zero-rated (lubricating oils are also subject to GST at the standard rate).

  • Sanitary napkins, tampons, menstrual cups and other such products.

Estimated Maldives GST Paid by Commodity and Category of Taxpayer

article image

TEs in the GST, by Goods and Services

article image

Corresponds with “medical appliances, precision and optical instruments, watches and clocks” in the SUT.

There are interactions that are captured when all interactions are removed at once, but not captured when removing exemptions in isolation. Source: IMF staff estimates.

References

  • Canada, 2021, Report on Federal Tax Expenditures, Concepts, Estimates, and Evaluations, Department of Finance. Available at: taxexp-depfisc-21-eng.pdf (canada.ca)

    • Search Google Scholar
    • Export Citation
  • IMF, 2015, Options for Low Income Countries’ Effective and Efficient Use of Tax Incentives for Investment. Washington DC.

  • IMF, 2016, Fiscal Policies for Innovation and Growth, Fiscal Monitor, April. Washington DC.

  • IMF-OECD, 2020, Chile—Assessment of Tax Expenditures and Corrective Taxes, Technical Assistance Report.

  • Heady, C. and Mansour, M., 2019, Tax Expenditure Reporting and Its Use in Fiscal Management: A Guide for Developing Economies. International Monetary FundHow to Note, Washington DC.

    • Search Google Scholar
    • Export Citation
  • Hebous, S., and Burns, L., 2020, Maldives—Income Taxation: Recent Progress and Prospective Improvement. Technical Assistance Report, IMF. Washington DC.

    • Search Google Scholar
    • Export Citation
  • Hebous, S., Burns, L., and Norregaard, J., 2019, Maldives—Reform Options to Strengthen Tax Policy. Technical Assistance Report, IMF. Washington DC.

    • Search Google Scholar
    • Export Citation
  • Organisation for Economic Co-operation and Development. 2010. “Tax Expenditures in OECD Countries.” Paris.

  • Wen, J-F, 2020, Temporary Investment Incentives, IMF Special Series on COVID, Washington DC.

2

Earnings between MVR 720,000 and MVR 1.2 million are taxed at 5.5 percent; earnings between MVR 1.2 million and MVR 1.8 million are taxed at eight percent; and earnings between MVR 1.8 million and MVR 2.4 million are taxed at 12 percent.

3

For example, for a taxable income of MVR 2,000,000 (before the general deduction), the tax payment under the BPT is MVR 225,000 (1.5 million × 0.15) implying an average tax rate of 11.25 percent, whereas the tax payment under the PIT would be MVR 98,400 (i.e., the average tax rate would be 4.92 percent).

4

Thus, the Benchmark includes the rules that limit interest deductions; fix rental income deduction at 20 percent of rental income; limit head office expenses deduction, and limit on non-monetary remuneration deduction.

5

A resident company registered under the Companies Act (which was repealed) was liable for tax at the lower rate of 5 percent if the only income derived by the company are certain specified classes of foreign income. These include income from a business carried on wholly outside the Maldives and Income from any immovable property located outside the Maldives.

6

According to the authorities, Zakat-al-Mal is not fully enforced.

7

Data on the details of bank’s tax filings are limited but included where possible. Information on taxable income and tax payments was available for all banks.

8

The tax return data required a minimal cleaning procedure, which was discussed during the workshop and written documentation was provided to workshop participants.

10

There are many firms that are put into a loss position with full expensing and those losses would carryforward to subsequent years, potentially reducing BPT revenue for multiple years.

11

Ideally, one should separate the provision of such services between private and public provisions. The taxation of this private provision of services would be part of the benchmark while the GST exemption would constitute a deviation from the Benchmark and hence a TE.

12

The detailed threshold provisions are laid out in Chapter 10 (Registration) of the GST Act.

13

The Source of the data is the National Bureau of Statistics. Work is currently ongoing to produce the SUTs for 2017.

14

The mission obtained these ETRs from estimates by the Asian Development Bank (ADB).

15

These are exports in the SUTs and they are subject to the tourism GST in the Maldives.

16

Note that the data in the Use Table are expressed in terms of purchaser price (i.e., including the GST). Therefore, estimating GST revenues requires dividing the expenditures in the SUTs by (1 + ETR). For example, if household expenditures on the commodity “Accommodation, food and beverage services” are equal to MVR 1,191 million and the estimated ETR on this commodity is equal to 6 percent, GST revenues from this category will be equal to 1,191 ×(0.06/1.06). See also Box 1.

Maldives: Technical Assistance Report—Estimating Tax Expenditures
Author: International Monetary Fund. Fiscal Affairs Dept.