WORLD BANK WORKING PAPER Fiscal and distributional implications of VAT reforms in Zimbabwe Dhiraj Sharma Tawanda Chingozha Maya Goldman Indira Iyer Victor Steenbergen © 2025 International Bank for Reconstruction and Development / The World Bank 1818 H Street NW, Washington, DC 20433 Telephone: 202-473-1000; Internet: www.worldbank.org This work is a product of the staff of The World Bank with external contributions. The findings, interpretations, and conclusions expressed in this work do not necessarily reflect the views of The World Bank, its Board of Executive Directors, or the governments they represent. The World Bank does not guarantee the accuracy of the data included in this work. The boundaries, colors, denominations, and other information shown on any map in this work do not imply any judgment on the part of The World Bank concerning the legal status of any territory or the endorsement or acceptance of such boundaries. Nothing herein shall constitute or be considered to be a limitation upon or waiver of the privileges and immunities of The World Bank, all of which are specifically reserved. Rights and Permissions Attribution—Please cite the work as follows: Sharma, Dhiraj; Chingozha, Tawanda; Goldman, Maya; Iyer, Indira and Steenbergen, Victor (2025). “Fiscal and distributional implications of VAT reforms in Zimbabwe”. The World Bank, Washington, DC: World Bank. Translations—If you create a translation of this work, please add the following disclaimer along with the attribution: This translation was not created by The World Bank and should not be considered an official World Bank translation. The World Bank shall not be liable for any content or error in this translation. Adaptations—If you create an adaptation of this work, please add the following disclaimer along with the attribution: This is an adaptation of an original work by The World Bank. Views and opinions expressed in the adaptation are the sole responsibility of the author or authors of the adaptation and are not endorsed by The World Bank. Third-party content—The World Bank does not necessarily own each component of the content contained within the work. The World Bank, therefore, does not warrant that the use of any third- party-owned individual component or part contained in the work will not infringe on the rights of those third parties. The risk of claims resulting from such infringement rests solely with you. If you wish to reuse a component of the work, it is your responsibility to determine whether permission is needed for that reuse and to obtain permission from the copyright owner. Examples of components can include, but are not limited to, tables, figures, or images. All queries on rights and licenses should be addressed to World Bank Publications, The World Bank Group, 1818 H Street NW, Washington, DC 20433, USA; e-mail: pubrights@worldbank.org. 2 Fiscal and distributional implications of VAT reforms in Zimbabwe1 Dhiraj Sharma Tawanda Chingozha Maya Goldman Indira Iyer Victor Steenbergen February 2025 ABSTRACT Improving domestic revenue mobilization extremely important for Zimbabwe to create the fiscal space to absorb quasi-fiscal expenditures and support macroeconomic stability. In November 2023, Zimbabwe announced measures to raise additional tax revenue. This included limiting VAT zero ratings to exports only, and VAT exemptions to a small number of essential items. This paper carries out a VAT tax gap analysis and uses a partial fiscal incidence analysis framework to analyze the welfare and distributional consequences of the reforms. The change announced in the 2024 budget is expected to increase the VAT revenue by 0.88 percent of GDP. But it also increases the poverty headcount by 1.4 percentage points and inequality by 0.14 points. This pattern is consistent with international evidence. Therefore, any VAT reform must be accompanied by a compensation mechanism with horizontal expansion, i.e., a broader coverage of less well-off households. Focusing only on the current beneficiaries of the cash transfer program is ineffective in restoring the welfare level because of its minimal coverage level of the total population and the poor. This can be done with a fraction of the new VAT collections from the proposed changes. Zimbabwe does not have a unified social registry and a criterion to target the least well-off households. Thus, instituting a compensation mechanism requires significant investments in establishing a nationwide social registry and developing a targeting mechanism to target social cash transfers effectively. Keywords: Value Added tax, VAT exemptions, distributional analysis, cash transfers, microsimulation 1 This paper was produced as a background paper for the World Bank’s “Zimbabwe Public Finance Review: Anchoring Macroeconomic Stability through Fiscal Policy” (2025). Special thanks to Rinku Murgai (Practice Manager, EAEPV), Pierella Paci (Practice Manager, EAEPV), and Abha Prasad (Practice Manager, EAEM1) for their helpful comments and support. 3 I. Introduction Taxes and transfers can be powerful tools to affect the post-market distribution of income to reduce poverty and inequality in a country. One approach commonly used to do so is to offer lower value-added tax (VAT) rates or VAT exemptions on items thought to constitute a larger share of total household consumption of the poorer households. As such, many countries use standard rates for consumer durables and luxury items (e.g. air conditioning, automobiles), while basic and essential items, such as food, sugar, and medicine, are exempted or zero-rated. Despite their widespread use, there is evidence that VAT is an inefficient way to redistribute resources. For example, a study from six low—and middle-income countries finds that preferential VAT has a poverty- reducing effect, but it is not well-targeted toward poor households. A more efficient way to reduce poverty and inequality is to expand the VAT base and redirect the increased revenues to poverty-targeted or universal cash transfer programs (Warwick et al., 2022). In November 2023, Zimbabwe announced various measures to raise additional tax revenue. This included a reform to limit VAT zero ratings to exports only, and VAT exemptions to a small number of essential items. Reducing the number of items eligible for VAT exemptions would increase the tax base and revenue. However, exemptions are often justified in the first place on the grounds that they make VAT more progressive, or that they help achieve a more equitable post-tax income distribution. So, it is important to understand the welfare and distributional implications of changes to the VAT structure. This also allows the possibility to adopt compensatory mechanisms, such as cash transfers that redistribute some of the gains of tax expenditures to low-income households to compensate for the “losses” due to the removal of VAT exemptions. This may help ensure that fiscal reforms achieve domestic revenue mobilization while ensuring the reforms are equitable and poverty-reducing. In this paper, we combine a VAT gap analysis with a partial fiscal incidence analysis framework to assess the fiscal and distributional impact of the reforms announced in the 2024 budget. The subsequent sections describe the existing VAT structure and the proposed reforms (section II), the methodology and data (section III), and the main findings (section IV). Finally, section V concludes with policy recommendations. II. VAT structure in Zimbabwe and the proposed changes History of VAT in Zimbabwe VAT is a consumption or destination-based indirect tax that is charged on consumption or supply of taxable commodities and services (ZIMRA, 2023; Mpofu, 2022). Globally, it accounts for a fifth of government revenues (Keen & Lockwood, 2007). In Zimbabwe, it is the largest contributor to tax revenue since 2009, having overtaken Pay-As-You-Earn (PAYE) which was the country’s primary revenue source (Maveneka, 2011). VAT (together with Import VAT, surtax, as well as Customs and Excise Duty) constitutes an indirect tax charged on certain groups contingent on consumption. It is charged at different stages of the value chain (Munyoro et al., 2018). It is charged on transactions rather than income or profit, as well as on the importation of services and commodities (ZIMRA, 2023). In Zimbabwe, VAT is regulated by the country’s Value Added Tax Act (Chapter 23:12) and the VAT Regulations SI 273 of 2003. VAT is levied on a transaction basis, because of which the responsibility to charge the tax on price lies upon a registered business operator whether they turn a profit or not (Munyoro et al., 2018; Mpofu, 2021). 4 Today’s tax system can be traced back to the IMF and World Bank-led reforms of the 1990s that sought to create a politically independent revenue authority and replace sales tax with VAT to curb tax evasion (Zhou & Madhikeni, 2013). A fairer tax architecture that is more effective in mobilizing revenue and less complex to administer was another motivation behind replacing sales tax (Munyoro et al., 2018). The tax reforms of the 1990s also recommended segmenting the tax-payer base into micro, small, medium, and large groups (Zhou & Madhikeni, 2013) to enable tax administration to respond better to their different needs. As of 2023, the turnover thresholds for the different segments of the VAT categories/groups are shown in Table 1. Failure to register for VAT is an offense under the VAT Act (Chapter 12:12). If an operator failed to register for VAT, yet their annual sales turnover exceeded a minimum threshold (USD40,000 before 2024, and USD25,000 thereafter), the Commissioner may compulsorily register the operator, and they would be obligated to pay any tax due as well as interest and penalties (ZIMRA, 2023). There are other economic activities for which operators are not obliged to register for VAT. This list includes operators trading exclusively in exempt supplies, persons engaging in private/recreational pursuits, any service rendered by an employee to his employer, and small traders whose taxable annual turnover is less than USD25,000. Table 1: VAT Categories and Turnover Thresholds Category Annual Turnover Registration Threshold (ZWL & USD) Group A minimum turnover level is set by the Commissioner from time to Voluntary Registration Micro time Compulsory Registration USD25,000 or ZWL Equivalent Small Category C Threshold USD240,000 or ZWL110,000,000 Large Category D Threshold USD120,000 or ZWL50,000,000 Medium Source: ZIMRA (2024) Zimbabwe administers four different types of VAT rates. These are a 15% standard rate, 5% levied on export of raw hides, 20% on un-beneficiated chrome2 and 0% for specified goods and services as indicated in the VAT Regulations. The standard 15% rate applies to all goods and services unless specified as zero-rated or VAT exempt (PWC, 2023). The zero rate previously applied to Zimbabwean goods exported to an address outside the country, as well as basic food commodities listed in the VAT regulations. A registered operator’s tax obligation is the difference between output tax on sales and input tax on purchases/input materials, wherein ZIMRA is obligated to refund the operator if output tax is less than input tax (ZIMRA, 2023). VAT can directly enhance government revenue, but its indirect benefit is that it enhances the efficiency of the overall tax administration system (Keen & Lockwood, 2007). Within Zimbabwe’s context, the collection of VAT through mandatory use of devices that record taxable transactions to minimize losses may be a good example of tax administration system enhancement (Maveneka, 2011). Beginning on 1 April 2017, ZIMRA introduced a withholding tax whereby registered and appointed VAT operators are required to deduct two-thirds of the output tax from any amount paid to another registered operator who is specified to be liable to VAT withholding tax (Grant Thornton, 2022). Specified operators then claim the withheld two-thirds component on their VAT returns (Grant Thornton, 2022). In 2024, an additional 5 percent withholding tax was also introduced on any transaction with non-VAT registered companies (i.e., informal sector). Table 2 shows a timeline of selected VAT changes over the past few years. 2A 20% tax on unbeneficiated chrome was introduced on 1 August 2010 (PWC, 2015). Unbeneficiated chrome is chrome ore that has not been crushed, milled, washed, and smelted into pellets or ingots (PWC, 2015); hence, the tax encourages mining entities to set up refineries locally. 5 Table 2: Timeline of changes to the VAT statute over the past few years Statutory Date Description of change Instrument Source (SI) Standard rating of accommodation services provided to Jan 2015 10/2015 ZEPARU (2015) tourists Exempted certain commodities that were previously zero- Feb 2016 9/2016 Chronicle (2016) rated to reduce the number of zero-rated goods/services (a) Exemption of the supply of radiation protection services by the Radiation Protection Authority of Zimbabwe for 2011- 2015. (b) Also, zero rating of "the supply of pipeline Zimbabwean Dec 2016 transportation, storage and handling services for the purposes 154/2016 Government of delivering fuel through the pipeline with effect from 2nd Gazzete (2016) February 2009" and (c) zero-rating gold supply to Fidelity Printers & Refiners Reinstated a 15% on basic commodities such as rice, maize Feb 2017 groats and meal, meat and fish products, potatoes, mahewu 20/2017 FAO (2017) (cereal based drink) and margarine FAO (2017), Veritas Feb 2017 Repealed SI 20/2017 due to inflationary concerns 26A/2017 (2017) Exempted ancillary services provided by National Mar 2018 30/2018 Kubatana (2019) Pharmaceutical Company (NatPharm) July 2018 Exempted services offered by messenger of court 138/2018 Kubatana (2019) Zimbabwean Exemption of goods and services provided/imported by Dec 2019 284/2019 Government statutory bodies Gazzete (2019) Ndhlovu et al., Jan 2020 Reduction of VAT from 15% to 14.5% effective 1 January 2020 (2022), Grant Thornton (2022) The standard rate was changed to 15% with effect from 1 Jan 2023 PWC (2023) January 2023 Removal of most VAT zero-rating and some VAT exemption of goods and services. Lowered VAT threshold to USD 25,000. MoF (2023) and Jan 2024 15/2024 Introduced a 5% withholding tax on any transaction with non- ZIMRA (2024) VAT registered companies (i.e., informal sector). 2024 changes to VAT structure To increase domestic revenue mobilization, the Government of Zimbabwe announced in the 2024 budget that most basic commodities that were previously zero-rated would be shifted to exempt3 status, and the rest of them (for example petroleum jelly) to standard-rated status (Table 2). Some items such as fish (and seafood) and meat that were previously on the exempted list would also be standard rated. Following the pronouncements in the budget and subsequent publishing of Statutory Instrument (SI) 15 of 2024, basic commodities previously zero-rated are now exempt, and the zero rate now mostly applies to exports. As a result, items like rice, sugar, fish (and seafood) and meat are now standard rated, as well as petroleum jelly and bath soap. Examples of VAT-exempt supplies include financial services, water and electricity for 3 For a “zero-rated” good, the government does not tax its sale, but allows credits for the value-added tax paid on inputs. If a good or business is “exempt,” the government does not tax the sale of the good, but producers cannot claim a credit for the VAT they pay on inputs to produce it. As such, the tax burden for zero-rated items is lower for businesses than for exempt (EXEMPT?) items (and tax expenditures to the government are higher). 6 domestic use, and municipal rates (ZIMRA, 2023). Children’s clothing, printed books, and almost all food items were zero-rated upon entry into Zimbabwe (ZIMRA, 2023). Please refer to Table A1 in the Annex for classification of items, and their breakdown by standard, exempted and zero-rated category before, and after 2024. Table 3: Change in the VAT schedule from the 2024 budget (percentage of items in standard, exempt and zero-rated category) Original VAT Schedule 2024 VAT Schedule Change Standard Zero Standard Zero Standard Zero Items Exempted Exempted Exempted Rated Rated Rated Rated Rated Rated Food items 31 52 17 45 55 0 +14 +3 -17 Other items 67 22 10 71 29 0 +4 +6 -10 Total 54 33 13 62 38 0 +8 +5 -13 Limiting the number of VAT-exempt or zero-rated commodities will likely boost the fiscus by increasing the efficiency and neutrality of VAT tax collection (Warwick et al., 2022). Yet, it can also hurt low-income households, thereby undermining the government’s objectives of equitable and progressive tax policy. As such, Zimbabwe may need to consider compensatory mechanisms to complement the removal of VAT exemptions. This may help ensure that fiscal reforms achieve domestic revenue mobilization while ensuring the reforms are equitable and poverty-reducing. III. Data and methodology for modeling the impact of VAT reform Data The analysis uses several data sources. The information on household consumption comes from the 2017 Poverty, Income, Consumption, Expenditure Survey (PICES). The survey collected detailed household consumption and income information from more than 30,000 households. The food consumption data was captured using a 30-day diary method, in which households filled out a diary with daily entries on items consumed and the source of consumption: market purchase, own production, in-kind payments, gifts, or social transfers. Expenditure on non-food items was captured using the 30-day recall method where households reported total expenditure over the previous month on various non-food items. The official measure of welfare in Zimbabwe is per capita consumption. Consumption is theoretically a better measure of long-term standard of living settings such as Zimbabwe, where income is erratic and informal. Without a more recent Supply and Use Table (SUT), we use the table developed for 2012. The SUT consists of two interlinked tables: the supply and use tables. The supply table shows the availability of goods and services in a country over a period, including domestic production and imports. The use table shows how goods and services are used in the country as intermediate or final goods. The 2012 Zimbabwe SUT provides information on the total supply of products for 120 product categories and their use as final consumption, exports, or intermediate consumption in 36 sectors. Information on VAT collection and structure is obtained from disparate sources. Information on total and sectoral VAT collection comes from Zimbabwe’s Ministry of Finance and ZimRA. We mainly rely on the VAT Act and VAT Regulations to understand the rates for different supplies. Where the appropriate tax rate is 7 not obvious in both the VAT Act and Regulations, we rely on alternative sources such as the websites hosted by ZIMRA and international accounting firms such as PwC and Deloitte. We use the informality database developed by Bachas et al. (2020) for 32 low- and middle-income countries to impute the share of informal purchases in Zimbabwe. The database has the share of consumption from informal sources for each decile disaggregated up to the fourth consumption subcategory.4 For the estimate of aggregate informality for Zimbabwe, we use the informal economy database developed by the World Bank’s Prospects Group.5 Methodology VAT tax gap analysis The VAT tax gap is calculated using the 2012 Supply-Use Tables (SUT), which help disaggregate the overall tax gap into policy and compliance gaps. We adjusted the SUT table to analyze the VAT base for taxable purposes (see Box 1). Using the adjusted SUT table, the VAT gap is calculated as the difference between the potential revenue of the underlying economic tax base and actual revenue (Figure 1). This gap is decomposed into two main components: the impact of policy choices (policy gap) and noncompliance (compliance gap).6 Figure 1: Calculating the VAT gap, policy gap and compliance gap The is calculated using the standard rate of 15% in 2012 and 14.5% in 2022. This gives the outside envelope for total tax collections. We then consider the 2022 VAT policy with zero-rated commodities and exemptions. In the absence of unincorporated enterprises and firm-level survey data, we assume that all firms in the formal sector are registered. While presumptive taxes apply only to a small proportion of businesses at varying rates in Zimbabwe, we have assumed that 3% of firms pay presumptive taxes at an average rate of 3%. The VAT policy is calculated as follows: 4 For example, bread and cereals is the fourth level of disaggregation, after food (level 3), food and non-alcoholic beverages (level 2), and all products (level 1). 5 Informal Economy Database: https://www.worldbank.org/en/research/brief/informal-economy-database 6 Hutton, Eric. 2017. The Revenue Administration-Gap Analysis Program: Model and Methodology for Value-Added Tax Gap Estimation. IMF, Fiscal Affairs Department. 8 = ∗ ((1 − % ) ∗ ℎ + % ∗ 0.03) The difference between the total VAT potential using the standard rate and the VAT potential under the current VAT policy is called the VAT policy gap and is given by: = − The would be the maximum that can be collected under existing VAT law. Hence, if the actual VAT collections are below , this would be due to a lack of compliance. The VAT compliance gap is hence given by: = − We estimate the VAT policy gap in two cases—the first is the pre-2024 situation (baseline), and the second is an approximation accounting for the 2024 reforms. Box 1: Adjusting the Supply-Use Table to analyze VAT base Supply Table and Use Table are Commodity X Sector (Row X Column) matrices, or alternatively referred to as Product X Industry matrices, of equal dimensions with different entries. In general, the SUT is represented as an matrix, where are commodities and sectors. In Zimbabwe, the SUT table has 120 commodities and 36 sectors. In the Supply Table, entries across columns show the value of the respective commodities by kind of supplier, distinguishing the domestic supply from foreign supply (imports). On the other hand, entries across the column in a Use Table show use of the respective commodities or intermediate consumption by sectors, final consumption, gross capital formation and exports. The supply use equation for any given product in an economy is mathematically expressed as: Output + Imports = Intermediate Consumption (IC) + Final Consumption [Government (GFCE) and Private (PFCE)] + Gross Capital formation (GCF) + Exports To maintain the mathematical identity, due adjustments for price differentials are required. Since output is compiled at basic prices (BP), net taxes on products need to be added on the left side of the equation. Accordingly, the above equation has to be re-written as: Output - Intermediate consumption + Taxes on products – Subsidies on products + Trade and Transport Margin (TTM) = Final consumption (government and private) + Gross capital formation [fixed (GFCF), changes in stocks (CIS) and valuables) + Exports – Imports The VAT gap can be computed using the Supply and Use Tables (SUT) using the above identity. Several adjustments are made to the SUT table to compute the VAT gap and use the above identity. For instance, we first need to arrive at the total available output in the economy by distributing imports and trade margins across the various sectors. As the Use Table is at the producer’s prices, taxes must be deducted from each sector to arrive at the tax-exclusive Use Value. We would also need to distribute the capital (GFCF) across intermediate and final use. Finally, we need to calculate the total ITC, a critical input to this computation. ITC is assumed to be collected for intermediate use. It is assumed that the ITC given back to the producer reduces the taxable base as production costs come down. These adjustments are then made to the Supple Table to compute the VAT Base. 9 Distributional analysis (fiscal incidence approach) We use a fiscal incidence approach to measure the VAT burden in Zimbabwe, and the impact of current and potential reforms across the distribution of households. The process can be disaggregated into 4 steps: i) Prepare a dataset of household consumption, ii) Generate and augment the Input-output table, iii) Impute the share of informal consumption by each decile and COICOP category, and iv) Simulate direct and indirect VAT payments for each consumption item and each household. Household consumption Preparing the datasets for the analysis involves several steps. We begin by preparing a dataset of household consumption from the PICES 2017, merging the food and non-food datasets together, removing outlier households, and dropping certain consumption items that do not incur VAT (such as loan amounts and interest rates). We also split the use-value of durables across IO sectors and COICOP codes using the share of total durables in each IO sector and COICOP category as a proxy.7 Next, we use Zimbabwe’s VAT Act and VAT Regulations to map the consumption items in the household survey to the statutory VAT rate and status (standard-rated, exempt, or zero-rated).8 Household Budget Surveys (HBSs) typically do not cover the universal set of consumption items/services. Certain goods and services (for example, minerals and other goods used in intermediate production) may or may not be partially/fully captured in household consumption. To ensure that we account for the universal consumption set as much as possible, we augment the list of consumption items in PICES 2017 with additional supplies based on the listing of the imports and apply the appropriate VAT rate to this augmented list. Next, we separate the share of market purchases from total consumption, as VAT is only incurred on market transactions. At the sub-group (COICOP category) level, the PICES 2017 data provides information on the value of consumption sourced from the market, received as transfers or gifts, and from own production. Based on this information, we calculate market and own production shares at the COICOP subgroup level and apply these to individual PICES 2017 consumption items that fall under the same category. Finally, we also map the consumption items in the household survey to the SUT sectors. We assign each supply line (consumption item) in the augmented list of items (obtained from PICES 2017, VAT Regulations, and import data) to the 120 SUT sectors to enable estimation of the tax gap. Again, for some commodities, it is not obvious which sector they belong to, so we use our best judgment. Input-output table An input-output table summarizes the interlinkages between the different industries in the economy. We use it to determine how unreclaimed VAT on inputs affects the final price of a good. We prepare an Input- Output (IO) table by combining some sectors in the Supply table and adding others until the table is square. 7 The use value is provided in the raw dataset as one aggregate value per household, but for our purposes, we need it to be disaggregated. 8 The VAT Regulations provide information on supplies that are exempt and zero-rated, hence we assume that any supplies not specified in them are standard rated (see Table A1 in the Annex). 10 After this step, the supply table records output items for each industry in their industry (i.e., there are no secondary outputs classified under other industries). This simplifies the process, and we can treat the supply table as the IO table directly, without having to transform any secondary outputs. We then calculate the technical coefficients (the cost structure of each industry’s outputs), by dividing the total value of each input commodity as a share of the total production value in each industry. The final step is to “augment” the IO table by disaggregating any sector with a combination of exempt and non-exempt items to treat these differently in our estimation of the indirect VAT rates. Informality Using indirect taxes as a fiscal redistribution tool requires effective tax administration. Vendors may not pay VAT for various reasons, including tax evasion, exemptions, and policy thresholds designed to protect small enterprises from an overly heavy administrative burden. In low- and medium-income countries, an important share of purchases is made with these “informal” vendors that do not pay VAT at the final point of sale (direct VAT), for example, unregistered retail outlets, street vendors, farmgate buyers, and so on. These sellers are also unable to reclaim VAT on their inputs (indirect VAT). Indirect VAT can cascade to the final point of sale, meaning that even exempt or informal items incur VAT. Furthermore, standard-rated goods can incur a rate of VAT that is higher than the statutory rate if there is unclaimed VAT included on exempt inputs into those goods. Generally, the place of purchase is used as a proxy for determining whether a seller is registered for VAT or not9, but this is not available in the 2017 PICES. Instead, we impute the variation in informal consumption by decile and product based on a simple average of nine comparator countries, while forcing the aggregate level to match Zimbabwe’s national estimates.10 The comparator countries are chosen because their aggregate informal share of GDP is within 15 percentage points of Zimbabwe’s aggregate informality rate of 60%.11 Macro-micro validation We carry out macro-micro validation to check how much of the consumption and VAT is captured in the survey relative to national accounts and administrative data. This is shown in Table 4Error! Reference source not found. The total VAT receipt in 2017 was $1092 million USD. According to the National Accounts, total final consumption, including that of non-profit institutions, was $17,491 million. The total VAT receipt from the survey totals up to $175 million, or 16% of the total VAT. Total household consumption from the survey is $11,331 million, or 65% of household consumption in National Accounts. Thus, the effective VAT rates, according to the administrative and survey data, are 6.2 and 1.5 percent, respectively. Table 4: Household consumption and VAT revenues in 2017 Survey: Admin Admin/National Accounts Survey Ratio VAT revenue 1,092 174.8 16.0% Final consumption: Household and NPISH 17,491 11,331.0 64.8% 9 For instance, supermarkets are assumed to be registered for VAT but not street vendors or small stalls. 10 In the Zimbabwe survey, the only available information is whether an item was purchased in an “urban outlet” or “non-urban outlet.” This does not provide the level of granularity needed to differentiate between formal and informal purchases. 11 The countries and the year of household income/expenditure survey are as follows: Montenegro (2009), Peru (2017), Sao Tome (2010), Congo (2005), Comoros (2013), Senegal (2008), Chad (2003), Cameroon (2014), and Burkina Faso (2009). 11 Effective rate 6.2% 1.5% 24.2% The most likely explanation for the difference in effective VAT rate estimated from the micro and macro data sources is a significant contribution of sectors like mining and quarrying and public administration to total VAT. Household surveys do not capture those sources, hence total VAT receipts are underestimated. This discrepancy may prejudice the robustness of our estimates at the national level. However, we are interested in the share of VAT paid by households, so we are confident in the robustness of the estimates relative to the size of household consumption. Simulating potential compensation scenarios Finally, we combine the results of the VAT tax gap analysis and the fiscal incidence analysis to analyze the fiscal and distributional effects of redistributing some or all the tax gains from removing VAT exemptions under different scenarios. We consider the following three scenarios: • Scenario 1 (Full re-distribution): All the additional revenue collected by removing exemptions agreed in the 2024 budget reforms are redistributed to the bottom 40 percent of households. • Scenario 2: (Compensating bottom 40 percent): Compensating the bottom 40 percent for all consumption losses from removing VAT exemptions. This could lead to potential fiscal savings while avoiding negative impacts on poverty and equity. • Scenario 3 (Compensating current cash transfer recipients): The third scenario considers the poverty and equity effects of compensating the losses to current cash transfer recipients. It contrasts the previous two scenarios by working through existing administrative systems (and highlighting their limited scope to target low-income households). We estimate the poverty and distributional impact of the compensation scenarios using the fiscal incidence approach (relying on 2017 PICES data). We calculate the sum of all the VAT allocated observed (baseline) and the sum of all the VAT allocated in the reform scenario (removing all VAT tax exemptions). The difference between these two sums is the total potential gains from removing VAT tax exemptions. For scenario 1, we allocate this amount equally to the bottom 40 percent of households. For scenario 2, for each household in the bottom 40 percent, we calculate their VAT payment in the reform scenario, subtract their VAT payment in the baseline, and allocate the difference in the form of a transfer. In other words, we fully compensate the “losses” (in the form of higher VAT payments) of the bottom 40 percent of households. For scenario 3, we replicate scenario 2 but restrict the transfers to the current beneficiaries of the Harmonized Social Cash Transfer (HSCT) program. For each scenario, we estimate the new poverty headcount and poverty gap after the transfer. Finally, we also use the VAT tax gap analysis to estimate the potential fiscal gains for each compensation scenario. To align these to 2022-2023 figures, we assume that each household’s general consumption patterns and share of household consumption-to-VAT collection remained constant over time. We then rely on the ratio of VAT transferred-to-total VAT collected in each of the scenarios and multiply this with the total VAT numbers for 2022-2023 to estimate the overall cost of compensating low-income households for their losses in consumption from VAT reforms. 12 IV. Results VAT gap analysis Current VAT tax gap To estimate the total VAT tax gap, we first consider the c-efficiency ratio – which is a composite measure of VAT efficiency that captures both the design and governance aspects of tax collection. The c-efficiency ratio is the ratio of actual total revenue to the theoretical maximum revenue that could be collected from a perfectly enforced tax levied at a uniform standard rate on all consumption. In Zimbabwe, the c-efficiency of VAT has fallen from 0.37 in 2012 to 0.29 in 2022, indicating policy and compliance gaps. Using the c- efficiency of VAT collections, a “back-of-the-envelope” VAT gap (adjusted for informal sector) indicates that the VAT gap was 5.2% in 2012, and it has risen to 5.7% in 2022 (Figure 2a). A more detailed disaggregation of the tax gap is computed using Supply and Use Tables (SUT) for Zimbabwe and detailed commodity-wise VAT tax rates and exemptions. Using the SUT, we estimate the VAT tax gap in 2022 was 5.4%, which is comparable to the VAT tax gap of 5.7% using the c-efficiency method (Figure 2b). To consider the overall cost of tax expenditures, we decompose the tax gap into a policy and compliance gap. Using the VAT tax policy in 2022, under which products are standard-rated, zero-rated, or exempt, the weighted average effective tax rate (ETR) is found to be 6.4%, against the standard rate of 14.5%. From that, we estimate that the VAT policy gap is 4.1% of GDP, which would be the total gains to the treasury if all exemptions were removed and zero-rated goods commodities were taxed at the standard tax rate. The VAT compliance gap is 1.3% of GDP, which is the maximum that could (theoretically) be gained in tax revenue through improved tax administration measures.12 This suggests that the VAT policy accounts for over three-quarters of the total gap (76%) while the VAT gap due to low compliance is 24%. Annex 2 provides a sectoral decomposition of the VAT tax gap and sectoral effective VAT rates. Figure 2: VAT tax gap a. Macro Level: c-Efficiency and VAT Gap b. VAT Policy Gap, Compliance Gap and Total Gap, 2022 12 The estimation of the compliance gap keeps informality at current levels and requires compliance only from the formal firms. 13 VAT Policy Gap under the 2024 reforms VAT policy changes as reflected in the 2024 budget could result in a revenue gain of 0.88 percent of GDP. Limiting the number of VAT exempt or zero-rated commodities boosts revenue collection by increasing the efficiency and neutrality of VAT tax collection (Warwick et al., 2022). The 2024 Budget has proposed several policy changes that moves an additional 15% of commodities in the SUT that were earlier zero-rated or exempt into the standard rate category. Earlier, in January 2023, the VAT standard rate was also increased from 14.5% to 15%. The partial equilibrium impacts of moving more commodities into the standard rate bracket, called rate rationalization (RR) here, and then applying an increased standard rate of 15%, is found to result in the effective tax rate (ETR) increasing from 6.4% in 2022 to 7.8% in 2024 due to RR, and then further to 8% due to an increase in the standard rate13 (Figure 3a). An increase in the ETR also results in the VAT tax gap falling from 5.62% of GDP in 2022 to an estimated 5.36% in 2024. Significantly, the policy gap decreases from 4.3% of GDP to 3.42% of GDP over the same period – or 0.88 percent of GDP (from 77% of the total gap in 2022 to 66% in 2024) (Figure 3b).14 Figure 3: VAT policy changes a. Impact of VAT policy changes on ETR b. Impact of VAT policy changes on VAT Gap The policy changes in the proposed 2014 Budget also result in a greater share of value-added in the economy taxed at standard rates though the share of zero or exempt items remains almost the same. In comparison to the 2022 VAT policy, the proposed changes in the 2024 Budget result in commodities taxed at standard rates increasing from 36% of value added to 39% (Figure 4). The share of commodities that are exempt or zero-rated remains almost the same at close to one-fifth of the value added. Figure 4: Impact of VAT policy change on commodities taxed at standard rates and less than standard rates 13 The combined impact of policy changes results in an increase in the ETR by 1.6% (1.4% due to RR and 0.2% due to an increase in the standard rate). However, it is to be noted that the ETR in 2024 is still low at 8%, or at just over half of the standard rate at 15%, suggesting that a significant VAT policy gap still remains. 14 This a simulated impact of the VAT policy reform, which may not be fully realized because consumers may have a behavioral response to higher VAT rates by switching their consumption bundle or increasing purchases from the informal sector. 14 Distributional analysis (fiscal incidence approach) Simulating VAT payments Aggregated by decile, we see that the share of consumption that is informal is highest for the poorest households at 83% and is decreasing with income (Figure 5a).15 It drops to 42% for the richest decile. We therefore expect to see that informality provides a level of protection from the burden of VAT payments for the poor. Nonetheless, because wealthier households consume much more than poorer households, the share of the total benefits from informality is higher for wealthier households. Overall, 17% of the total reduction in VAT collections due to informality benefits the richest 10% of the population, and only 1% benefits the poorest 10% (Figure 5b). Figure 5: Informality shares and concentration a. Share of informal consumption, by decile b. Concentration of informality benefits, by decile 40 Informal share of consumption (%) 83 1 Share of total benefit (%) 79 78 76 75 74 72 35 69 64 30 42 25 2 20 15 3 10 1 2 3 4 5 6 7 8 9 10 5 Decile of disposable income - 4 Informality benefits Informal share Average Decile of disposable income Source: Authors’ estimates based on the PICES 2017, Bachas et al. (2020), and the World Bank’s Prospects Group (2024). Using the augmented input-output table, the exemption status of each sector, and an average VAT rate for each non-exempt sector, an indirect rate of VAT is estimated for each IO sector. Using the VAT rate for each 15 Measured here by the welfare aggregate. 15 item and the indirect VAT rate for each IO sector, a pre-VAT value of consumption is backed out, and total VAT payments are estimated. To make this estimation, the model assumes that formal goods incur both direct and indirect VAT, but informal goods incur only indirect VAT.16 There is a substantial difference between the raw consumption we observe in the survey, and total consumption according to the welfare aggregate. Without more information on how the welfare aggregate is constructed, VAT payments are adjusted for this difference by calculating VAT as a share of raw consumption and applying that share to the welfare aggregate to determine final VAT payments per household. Pre-reform scenario Concentration shares show the distribution of total taxes across different population groups such as income deciles. A tax is progressive if its concentration curve lies below the Lorenz curve for disposable income, i.e., the tax burden falls on the richer parts of the distribution relatively more than disposable income. In this case, the Kakwani index – the difference between the concentration coefficient of tax and the Gini index – will be positive. Figure 6 shows the Lorenz curve for welfare aggregate (disposable income) and the concentration curve for VAT. The Gini index for disposable income is 44.7. VAT is progressive in that the share of VAT paid by the richer segments of the population is higher than the share of income going to the households. The concentration coefficient (or the “quasi-Gini”) of the VAT is 59.9. Hence, the Kakwani index equals 15.2. Figure 6: Progressivity of VAT 100 90 80 70 60 50 % 40 30 20 10 - 0 1 2 3 4 5 6 7 8 9 10 Decile of disposable income Cumulative share of disposable income Cumulative share of VAT Source: Own calculations using PICES 2017. Despite VAT being progressive, it is still the case that a larger share of the benefits due to VAT exemptions goes to the relatively better-off households. Almost 44% of the total VAT exemptions accrue to the richest 16This implies that informal vendors would choose to buy from informal vendors, and formal vendors choose to buy from formal vendors. There are incentives for vendors to behave this way because if informal vendors buy from formal vendors, they are unable to reclaim the VAT on their inputs, and vice versa. 16 decile, whereas only about 10% goes to the bottom 40 percent of the population (Figure 7). This is driven by the top decile’s higher overall consumption level and possibly a relatively higher share of spending on food items and other goods with preferential rates. So, while VAT exemptions do benefit the poor and the bottom 40 percent of the population, they are an inefficient way to reach them. Figure 7: Distribution of exemption benefits 50 Share of total benefit (%) 40 30 20 10 - Exemption benefits 1 2 3 4 5 6 7 8 9 10 Decile of disposable income Source: Own calculations using PICES 2017. Next, we examine the implications of removing tax distortions on the incidence and concentration of VAT, as well as poverty and equity. We consider two scenarios for closing the policy gaps: the removal of all exemptions and zero-rating, and the VAT schedule announced in the 2024 budget. Removing all exemptions and zero-ratings The 2024 VAT schedule shifts the VAT concentration curve considerably “inwards,” which means the VAT becomes substantially less progressive (Figure 8a). Indeed, the concentration curves for the 2024 VAT schedule and pre-reform disposable income overlap almost perfectly. The Kakwani Index falls from 13.3 at baseline to 1.3 with the changes introduced in the 2024 budget (Figure 8b). VAT also becomes slightly less progressive with the removal of all exemptions or zero-ratings because exempted items, such as food, account for a higher share of the household budget of the poor. In this scenario, the Kakwani index is 11.6. Figure 8: Progressivity of VAT reforms a. Concentration curves [Baseline, with the 2024 VAT schedule, and with no exemptions or zero-rating] 17 100.0 90.0 Cumulative share of 80.0 disposable income 70.0 Cumulative share of VAT, 60.0 baseline 50.0 40.0 Cumulative share of VAT, no exemptions or zero- 30.0 rating 20.0 Cumulative share of VAT, 2024 schedule 10.0 0.0 0 2 4 6 8 10 b. Kakwani Index (with respect to disposable income) [Baseline, with the 2024 VAT schedule, and with no exemptions or zero-rating] VAT (baseline) 13.3 VAT (no exemptions or zero-rating) 11.6 VAT (2024 budget) 1.3 Source: Own calculations using PICES 2017. The incidence of VAT shows the size of total VAT payments relative to household consumption. A higher incidence of VAT impoverishes households because, without additional income, fewer resources are available for consumption. As Figure 9 shows, the incidence of VAT is increasing in consumption deciles. With the complete removal of exemptions and zero-rating, the incidence is higher for all households. For the poorest decile, it more than doubles, increasing from about 0.9% of total household consumption to 2.3% (Figure 9). Nevertheless, VAT continues to be progressive due to informal purchases. The 2024 schedule eliminates the progressivity of VAT as all deciles pay approximately the same tax as a share of their disposable income. The incidence of VAT jumps to 4.2 percent of disposable income for the poorest decile, while it is 4.5 percent for the top decile. This is consistent with the almost perfect overlap of the concentration curves of VAT and disposable income in Figure 6(a). 18 Figure 9: Incidence of VAT (Baseline, with the 2024 VAT schedule, and with no exemptions or zero-rating) 6 % of disposable income 5 4 3 2 1 0 1 3 5 7 9 Decile of disposable income Baseline No exemptions or zero-ratings 2024 budget Source: Own calculations using PICES 2017. Despite much of the benefits going to the richer households, VAT exemptions do reduce poverty headcount and poverty gap. Table 5 shows the marginal effects of VAT exemptions on the poverty headcount and gap at different poverty lines. With total disposable income (consumption aggregate) less VAT, the poverty headcount ratio in Zimbabwe in 2017 was 30.6%. Without the exemptions, the headcount would be 31.5%, or 0.9 percentage points higher. The results would be qualitative the same at higher poverty lines. The poverty gap, i.e., the distance between household consumption and poverty line as a share of the poverty line, too would be higher. Inequality would be marginally lower without the VAT exemptions. The 2024 VAT schedule has the most severe impact on household welfare. If poorer households are not compensated for the higher VAT payments through other means, such as cash transfers, the poverty headcount increases by almost 2 percentage points, putting about 320,000 people in poverty (Table 5). Table 5: Poverty and inequality impact of removal of VAT exemptions Food poverty line Lower poverty line Upper poverty line Gini Headcount Gap Headcount Gap Headcount Gap Disposable income 44.7 30.0 7.7 52.9 19.8 69.2 32.9 Disposable income less 44.5 30.6 8.0 53.4 20.1 69.7 33.3 VAT (with exemptions) Disposable income less VAT (with the 2024 VAT 44.6 32.4 7.2 51.6 19.0 70.8 32.1 schedule) Disposable income less VAT (with no exemptions 44.2 31.5 8.3 54.4 20.7 70.5 34.0 or zero-rating) 19 Note: The 2017 food, lower, and upper poverty lines are US$29.8, $45.6, and $66.1 respectively. Source: Own calculations using PICES 2017. Removing informality Next, we examine the marginal effects of the removal of informality. Figure 10 shows the concentration curves and Kakwani Index under different scenarios. If all purchases were perfectly taxed, the concentration coefficient of VAT would drop from 58.0 to 54.2 (and the Kakwani Index would drop from 13.3 to 9.5). With a 10-percentage point reduction in informality (from 60% to 50%), the concentration curve would go from 58.0 to 57.1. Thus, with the removal of informality, VAT becomes less progressive, but it would still have a redistributive effect as it would stay more progressive than disposable income. Figure 10: Progressivity of removal of informality a. Concentration curves 100 90 80 70 Cumulative share of 60 disposable income 50 Cumulative share of VAT 40 Cumulative share of VAT, 30 50% informality 20 Cumulative share of VAT, 10 no informality - 0 1 2 3 4 5 6 7 8 9 10 Decile of disposable income b. Kakwani Index (with respect to disposable income) VAT (status quo) 13.3 VAT (50% informality) 12.4 VAT (no informaility) 9.5 Kakwani Index (with respect to disposable income) Source: Own calculations using PICES 2017. 20 Figure 11 shows the incidence of VAT with partial and complete elimination of informality. Removing all informality doubles households' VAT payment (relative to disposable income) in the bottom four deciles. Nevertheless, in absolute terms, the additional VAT collected from the richer households will be higher because of their higher level of consumption. Figure 11: Incidence of VAT (with partial and complete removal of informality) 4.0 3.5 % of disposable income 3.0 2.5 2.0 1.5 1.0 0.5 - 0 1 2 3 4 5 6 7 8 9 10 Decile of disposable income VAT 50% informality VAT, no informality Source: Own calculations using PICES 2017. The marginal effects of removal or reduction of informality on poverty and inequality is shown in Table 6. In the baseline scenario, 60% of the Zimbabwean economy is assumed to be informal as reported in the World Bank Informal Economy Database.17 Our analysis shows that informal purchases shelter households from the full impact of the VAT. If the VAT were perfectly administered on all purchases, the poverty headcount at the food poverty line would increase by about 0.6 percentage points (from 30.6% to 31.3%) and the poverty gap by 0.2. The increases would be smaller under a partial reduction of informality to 50%. Table 6: Poverty and inequality impact of removal of informality Food poverty line Lower poverty line Upper poverty line Gini Headcount Gap Headcount Gap Headcount Gap Disposable income 44.7 30.0 7.7 52.9 19.8 69.2 32.9 Disposable income less VAT (baseline - 60% 44.5 30.6 8.0 53.4 20.1 69.7 33.3 informality) Disposable income less 44.5 30.7 8.0 53.5 20.2 69.8 33.4 VAT (with 50% informality) 17 https://www.worldbank.org/en/research/brief/informal-economy-databas2e. 21 Disposable income less 44.4 31.3 8.2 54.2 20.5 70.2 33.8 VAT (with no informality) Note: The 2017 food, lower, and upper poverty lines are US$29.8, $45.6, and $66.1 respectively. Source: Own calculations using PICES 2017. Removing all distortions Finally, we consider both the complete removal of exemptions and zero-rating and a perfect administration of VAT (Figure 12). This is admittedly unrealistic, but it is considered an extreme case to establish the “upper bound” impact of the reforms. The joint effect of removal of informality and exemptions is significant. The concentrative curve without distortions is significantly “inward” of the baseline. Indeed, the concentration curve is now almost perfectly superimposed on the Lorenz curve, with the Kakwani Index of -0.1, which implies the VAT is slightly less progressive than the original household consumption. Therefore, exemptions and informality drive much of the progressivity of the VAT. Figure 12: Progressivity of removal of exemptions and informality a. Concentration curves (baseline, and with no distortions) 100.0 90.0 80.0 70.0 Cumulative share of 60.0 disposable income 50.0 Cumulative share of VAT, baseline 40.0 Cumulative share of VAT, 30.0 no distortions 20.0 10.0 0.0 0 2 4 6 8 10 22 b. Kakwani Index (with respect to disposable income) VAT (baseline) 13.3 VAT (2024 budget) 1.3 VAT (no distortions) -0.1 Source: Authors’ calculations using PICES 2017. Consistent with Figure 12a, the incidence of VAT is almost uniform throughout the welfare distribution, ranging from 11.0% of disposable income (for the bottom decile) to 11.3 % (for deciles 8 and 9). Figure 13: Incidence of VAT (baseline, and with no distortions) 12 % of disposable income 10 8 6 4 2 0 1 3 5 7 9 Decile of disposable income Baseline VAT (no distortions) Source: Authors’ calculations using PICES 2017. If all exemptions are removed and VAT is administered perfectly, the poverty headcount at the food poverty line would increase from 30.6% at baseline to 36.4%, or about 6 percentage points, impoverishing about 856,000 people (Table 7).18 The poverty gap would also increase by about a third. The increase in poverty gap is due to new households falling into poverty and the drop in consumable income of the existing poor households. Table 7: Poverty and inequality impact of removal of informality and exemptions Food poverty line Lower poverty line Upper poverty line 18 Using the estimated population of 14.75 million in 2017. 23 Gini Headcount Gap Headcount Gap Headcount Gap Disposable income 44.7 30.0 7.7 52.9 19.8 69.2 32.9 Disposable income less 44.5 30.6 8.0 53.4 20.1 69.7 33.3 VAT Disposable income less 44.7 36.4 10.6 58.7 23.8 73.8 37.2 VAT (with no distortions) Note: The 2017 food, lower, and upper poverty lines are US$29.8, $45.6, and $66.1 respectively. Source: Authors’ calculations using PICES 2017. 2024 budget reforms with compensation scenarios To summarize the analysis so far, although current administration of VAT in Zimbabwe is progressive, much of the benefits due to VAT exemptions and lax tax administration from informal purchases accrues to the wealthier households because of higher total consumption. However, a removal of exemptions and a stricter tax administration will increase the poverty headcount and poverty gap because, in absence of alternative sources of income, a higher share of disposable income goes to VAT payments, hence less is available for consumption. Therefore, VAT reforms must be coupled with compensatory mechanisms to restore the welfare of the poor and near-poor households. This section considers three different compensatory mechanisms and evaluates their welfare effects. Scenario 1: Full distribution of additional revenue to the bottom 40 percent of the population This scenario considers what happens to poverty and equity if all the additional revenue collected through the removal of exemptions announced in the 2024 budget reforms is redistributed to the bottom 40 percent of households. By definition, this scenario would be revenue-neutral to the treasury as all the additional revenue collected would be redistributed. This is admittedly unrealistic as the fiscus would not redistribute all additional taxes. However, the scenario is considered to establish an upper bound of the potential poverty-reducing effects of the reforms. As estimated from the VAT tax gap analysis, a total 0.88% of GDP would be available from additional VAT collections for new financing of direct cash transfers. Assuming that all the extra collections are equally distributed to the poorest 40 percent of households, it would reduce poverty by almost 8 percentage points, from 30.6% to 22.7% (Figure 14). It also shows that the current setup of VAT exemptions and zero- rating is an inefficient way to reduce poverty and inequality in Zimbabwe. Scenario 2: Compensate the losses of the bottom 40 percent of the population In this scenario, we consider compensating the bottom 40 percent for all the consumption losses from removing VAT exemptions in the form of higher VAT payments. In doing so, we assume that household market income stays constant and that higher VAT payments lead to a lower household disposable income. This compensation arrangement could lead to fiscal savings while avoiding negative impacts on poverty and equity. In this case, the post-transfer food poverty rate is identical to the baseline level because everyone in the bottom 40 percent will receive a reimbursement equal to the “loss” from higher VAT payments (Figure 14). 24 Yet, this scenario would provide considerable revenue to the treasury, as estimates suggest that the bottom 40 percent of the population can be compensated with only 17% of additional VAT collected. This suggests it would require 0.15% of GDP to compensate low-income households, while the treasury would still gain an additional 0.73% of GDP. As such, there is an opportunity to raise the total amount of (net) tax revenue for the treasury while still compensating low-income households. This scenario is unrealistic as it assumes perfect knowledge of the additional VAT payment and perfect targeting of the poorest 40 percent of the population. Yet, it establishes that the benefits of the reforms (additional revenue to savings to the fiscus) are considerably higher than the costs (increase in poverty headcount and inequality). Scenario 3: Compensate the losses of the current beneficiaries of the Harmonized Social Cash Transfer (HSCT) program. The third scenario considers the effects of compensating losses to current recipients of the Harmonized Social Cash Transfer (HSCT) program. This scenario is considered to show the effects of vertical expansion using the current beneficiaries of the largest cash transfer program. HSCT's coverage is miniscule—it covers only 0.4% of the population. Further, even the limited coverage is not well targeted to the poorer segments The coverage among the poorest and second poorest quintiles is 0.8% and 0.6%, respectively (World Bank, 2020). In this scenario, the fiscal cost of compensation is low. It would require only 0.3% of additional VAT collected (or 0.026% of GDP) to compensate current HSCT recipients. However, compensating only the HSCT recipients makes no dent in the higher poverty headcount and inequality caused by the changes to the VAT in the 2024 budget (Figure 14). This illustrates the inefficiency and ineffectiveness of the current social protection system. Figure 14: Poverty and equity effects of compensatory mechanisms A. Impact on poverty reduction B. Impact on inequality (Gini points) (food poverty line) 1.8 1.8 0.14 0.14 Percentage points change 0 Gini points change -0.36 -7.9 -2.9 2024 Budget Scenario 1: Full Scenario 2: Full Scenario 3: Full 2024 Budget Scenario 1: Full Scenario 2: Full Scenario 3: Full VAT Reforms redistribution compensation compensation VAT Reforms redistribution compensation compensation of gains to of "losses" to of "losses" to of gains to of "losses" to of "losses" to bottom 40% bottom 40% current HSCT bottom 40% bottom 40% current HSCT recipients recipients Source: Authors’ calculations using PICES 2017. V. Conclusion This note carries out a VAT tax gap analysis and uses a partial fiscal incidence analysis framework to analyze the welfare and distributional consequences of VAT reforms in Zimbabwe. VAT is more progressive than disposable income, so the removal of exemptions or stricter enforcement almost always reduces inequality, though by a small magnitude. At the same time, we find that VAT exemptions and purchases from informal sources shelter households from the full costs of the VAT. Hence, a removal of exemptions or stricter tax enforcement without complementary compensatory schemes will increase poverty headcount and poverty gap. The changes announced in the 2024 budget is expected to increase the VAT 25 revenue by 0.88 percent of GDP. But it also increases the poverty headcount by 1.4 percentage points and inequality by 0.14 points. Therefore, any VAT reform must be accompanied by a compensation mechanism with horizontal expansion, i.e., a higher coverage of less well-off households. This can be done with a fraction of the new VAT collections from the proposed changes. Focusing only on the current beneficiaries of the cash transfer program is ineffective in restoring the welfare level because of its minimal coverage level of the total population and the poor. Zimbabwe does not have a unified social registry and a criterion to target the least well-off households. Thus, instituting a compensation mechanism requires significant investments in establishing a nationwide social registry and developing a targeting mechanism to target social cash transfers effectively. 26 References Bachas, P., Gadenne, L., & Jensen, A. (2020). Informality, consumption taxes and redistribution (No. w27429). National Bureau of Economic Research. Grant Thornton (2022). Indirect tax – Zimbabwe. Available online: https://www.grantthornton.global/en/insights/indirect-tax-guide/indirect-tax---Zimbabwe/ Madzivanyika, E. (2017). A diagnosis of the deficiencies in the Zimbabwean value added tax system. Public and Municipal Finance, 6(2), 16-26. Maveneka, L. (2011, January). What has Tax Got to do with Development: A Critical look at Zimbabwe’s Tax system. African Forum and Network on Debt and Development (AFRODAD). Mpofu, F. Y. (2022). Taxing the digital economy through consumption taxes (VAT) in African Countries: Possibilities, constraints and implications. International Journal of Financial Studies, 10(3), 65. Munyoro, G., Mupfumira, A., Langton, I. and Dube, F. (2018). The Significance of Tax in the Promotion of Small Enterprises in Zimbabwe: A Case Study of Harare. Africa Development and Resources Research Institute Journal, Ghanopera: Vol. 27, No. 3(4), Pp. 1-28, E-ISSN: 2343-6662, 28th February, 2018. Ndhlovu, F; Muchazviona, D; Manhimanzi, G. C; Nyawo, T; Magoba, C; Karuru, B & Hlatywayo, M. (2022). An Assessment of the Implications of Standard VAT Rate Reduction towards the Improvement of Domestic Revenue Mobilization in Zimbabwe: The Case of Pam Golding Properties. Journal of African Interdisciplinary Studies, 6(10), 102 – 125. PWC (2023). Zimbabwe: Overview. Available online: https://www.pwc.co.za/en/publications/vat-in- africa/zimbabwe-overview.html PWC. (2015). Managing VAT in Africa – a mining perspective. Available online: https://www.pwc.co.za/en/assets/pdf/managing-vat-in-africa-a-mining-perspective.pdf ZIMRA. (2024). VAT Registration. Available online: https://www.zimra.co.zw/domestic-taxes/vat/vat- registration ZIMRA (2023). Mechanics of VAT. Available online: https://www.zimra.co.zw/domestic- taxes/vat/mechanics-of-vat 27 Annex 1: Original and 2024 VAT schedule (standard, exempted and zero-rated) Table A1: VAT items – standard, exempted, and zero-rated Original VAT Schedule 2024 VAT Schedule Standar Zero Standard Zero Items d Rated Exempted Rated Rated Exempted Rated Food items 93 (31%) 155 (52%) 50 (17%) 135 (45%) 163 (55%) 0 (0%) Alcohol & narcotics 20 1 0 21 0 0 Dairy, eggs 0 1 16 0 17 0 Fish, seafood 0 11 0 11 0 0 Flour products, cereals 5 6 22 5 28 0 Food away from home 27 0 0 27 0 0 Fruits, nuts 13 27 0 13 27 0 Meat 0 24 3 27 0 0 Oils, fats 1 2 5 1 7 0 Pulses 2 2 0 2 2 0 Seasoning, prepared food 20 0 1 20 1 0 Sweets 0 8 2 1 9 0 Tea, coffee 1 1 0 1 1 0 Vegetables 3 56 1 5 55 0 Water, beverages 1 16 0 1 16 0 Other items 351 117 54 372 150 0 Clothing & footwear 84 0 0 84 0 0 Coal 1 0 0 1 0 0 Communication 10 0 0 10 0 0 Education 0 18 0 0 18 0 Electronic goods 17 0 0 17 0 0 Forestry and logging products 3 0 1 4 0 0 Furniture 16 0 0 16 0 0 Glass products 3 0 0 3 0 0 Health 0 21 7 0 28 0 Household equipment & 57 1 3 58 3 0 services Home appliances 18 0 0 18 0 0 Housing repair 2 1 0 2 1 0 Insurance & other services 5 23 0 3 25 0 Manufacturing 3 4 7 12 0 0 Mining 53 0 4 57 0 0 Personal care 29 8 3 31 9 0 Pet food 0 0 1 0 1 0 Plaster lime and cement 5 0 1 6 0 0 Recreation 9 1 16 9 17 0 Restaurants & accommodation 2 0 8 3 7 0 Rubber and plastics products 5 1 1 7 0 0 Tobacco 0 2 0 2 0 0 Transport 0 19 1 0 20 0 Transportation 27 0 0 27 0 0 Utilities 2 20 1 2 21 0 Grand Total 444 272 104 507 313 0 Source: World Bank analysis using ZIMRA reports, 2024 GoZ Budget, and Statutory Instrument 10A of 2024. Note: Blue indicates an increase in number of VAT items since 2024, red indicates a decline since 2024. 28 Annex 2: Effective VAT Rates Across Sectors In Zimbabwe, the manufacturing sector, the wholsale and retail trade sector and the real estate sectors account for almost two-thirds of the VAT gap. The VAT gap in the manufacturing sector is the highest at 2.04% of GDP, followed by the VAT gap in the wholesale and retail trade sector at 0.97% of GDP and then in real estate activities at 0.66% of GDP (Figure A1a). If we decompose the VAT gap into the policy and compliance gaps, the policy gap accounts for 65 to 70 percent of total VAT gap in the manufacturing and real-estate sectors, while the compliance gap at close to 90 percent of total VAT gap dominates in the wholesale and retail trade sector (Figure A1b). Figure A1: Sector-wise VAT GAP a. Sector-Wise VAT Gap as a Percent of GDP b. Sectoral Policy and Compliance Gaps In the manufacturing sector, four large sub-sectors account for close to two-thirds of the manufacturing sector gap (Figure A2). The VAT gap in the electrical equipment and machinery sub-sector at 0.41% of GDP accounts for one fifth of the total manufacturing sector VAT gap. The VAT gap is also significant for chemical and pharmaceutical products at 0.36% of GDP followed by the motor vehicles and communications equipment and the food, beverages and tobacco sub-sectors, with both at 0.26% of GDP. Figure A2: Incidence of VAT (with no exemptions or informality) a. Manufacturing Sub-sectors Gap as a b. VAT Gap in the Manufacturing Sub-sectors (as a Percent of Manufacturing VAT Gap, 2022 Percent of GDP, 2022) Source: Authors’ estimates based on GoZ SUT tables, MoF and ZimRA data, and ZIMRA reports. 29 A sectoral decomposition of value-added and effective VAT rates indicates that just four sectors – mining and quarrying, wholesale and retail trade, construction, and information, posts and telecommunications – that account for 36% of total value-added in 2022 are taxed at the standard rate of 14.5% of GDP in 2022 (Figure A3). The manufacturing sector that contributes to almost a fourth of value-added at 24% is taxed at an effective tax rate of 6.7%. Almost one-fifth of value-added at 18% that includes electricity, water, transportation and storage, real estate activities, public administration, education and health services, have an effective tax rate of zero percent. Figure A4 has further details on the effective tax rates across more granular sub-sectors of activities as in the SUT table. Figure A3: Share of value added in 2022 taxed at standard rates and less than standard rates 30 Figure A4: Effective VAT Rates Across Sectors Source: Authors’ estimates based on GoZ SUT tables, MoF and ZimRA data. 31