Tax Expenditure Study Report_Dec 2018.pdf

ZAMBIA TAX EXPENDITURE

STUDY

MINISTRY OF FINANCE, LUSAKA ZAMBIA

DECEMBER 27, 2018

DAVID M. MUKUNYA

ABBREVIATIONS AND ACRONYMS

CSO Central Statistical Office GoZ Government of Zambia K Kwacha (Zambia unit of currency)

RA-GAP IMF Revenue Administration Gap Analysis Program Model

TARC Tax Administration Research Centre TE Tax Expenditure

ZACCI

ZDA

Zambia Chamber of Commerce Zambia Development Agency

ZRA Zambia Revenue Authority

TABLE OF CONTENTS

EXECUTIVE SUMMARY

1. INTRODUCTION

1.1 Definition of Tax Expenditure

1.2 Justification for Reporting Tax Expenditures

2. METHODOLOGY

2.1 The Benchmark Tax System

2.2 Data Sources

2.3 Limitations

2.4 Corporate Income Tax (CIT)

2.5 Trade-Related Taxes (Customs and Excise Duties)

2.6 Value Added Tax

3. TAX EXPENDITURE ESTIMATES

3.1 CIT Expenditures

3.2 Trade-Related Tax Expenditures

3.3 Value Added Tax

3.4 Tax Expenditure in Relation to Employment

3.5 Tax Expenditure in Relation to Profitability

4. CONCLUSIONS AND POLICY RECOMMENDATIONS

SUGGESTED FRAMEWORK FOR REPORTING TAX EXPENDITURES

ANNEXES

APPENDIX: INVENTORY OF TAX INCENTIVES

REFERENCES

EXECUTIVE SUMMARY

Zambia witnessed macroeconomic challenges between 2014 and 2016, resulting in a decline in tax revenue as a percentage of GDP from 15.5 percent to 12.9 percent. Public debt rose from 35.6 percent to 60.5 percent of GDP. This unsustainable trend compelled the Government to take drastic measures that included reviewing its fiscal policy in a bid to increase domestic revenues while protecting social protection programs.

One area for government reform was tax expenditure, a critical integral component of the tax policy. Tax expenditure may be defined as potential revenue foregone by the government as a result of tax relief granted. The primary objective is to ease the tax burden on particular groups of taxpayers and to encourage investment in target areas and activities. The tax relief may be statutory or discretionary and can take the form of reduction in the tax rate, tax base or tax due.

Unlike direct budget spending, tax expenditure does not require annual legislative approval and is often open-ended. Its obscure nature means lack of transparency and accountability, often with unintended consequences.

The Zambia Ministry of Finance plans to institute a framework for reporting on tax expenditures periodically. This effort is intended to provide a holistic view of government spending and enable policymakers identify and estimate the cost of each provision in the tax structure, among others goals.

This study analyses data for 2015 and 2016, the years for which data were available, and covers the main tax types – corporate income tax, customs duty and value added tax. Ideally, it would be necessary to consider several tax years in order to determine the trend of tax expenditures. This report seeks to provide a building block for future reporting on tax expenditures.

According to the estimates, the total amount of revenue foregone was K 34.5 billion in 2016 and K 25.7 billion in 2015, representing 16 percent and 14 percent of GDP, respectively. Contextually, if tax expenditure and direct public spending were computed on the same basis, the amount foregone in 2016 would be equivalent to government expenditure on general public services, economic affairs function and defence that year.

The findings highlight the necessity for Government to rationalize tax expenditures. The current tax structure appears designed to address both systematic and firm-specific issues, rendering it too generous and redundant in several respects. Among the remedies, the authorities need to clearly describe the relief granted, state the objectives, have sunset and claw-back provisions, and design a clear framework for assessing the efficiency and effectiveness of the programs.

1. INTRODUCTION

Between 2014 and 2016, Zambia witnessed macroeconomic challenges, with real GDP growth slipping from 4.4 percent to 3.4 percent. Over this period, public debt as a percentage of GDP jumped from 35.6 percent to 60.5 percent, while tax revenue declined from 15.5 percent to 12.9 percent as a percentage of GDP. This was attributable to a combination of unfavourable exogenous and fiscal policy conditions. A decline in copper prices led to low export revenues, while poor rainfall reduced the supply of hydro-electricity electricity and agricultural output. In addition, the high volume of tax expenditures, in form of tax exemptions and reduced tax rates, further constrained Government’s ability to raise adequate revenue.

The unsustainable trend compelled the Government to take drastic measures that included phasing out regressive fuel and electricity subsidies, and scaling up expenditure on social protection programs. To increase domestic revenues, action was necessary to address widespread exemptions and broadening the VAT and income tax bases.

Furthermore, the Government launched the Economic Stabilization and Growth Program, aimed at restoring macroeconomic stability and creating conditions for sustained growth. The program placed heavy emphasis on enhancing resource mobilization, strengthening accountability and transparency in the use of public resources, and restoring budget credibility, among others.

A critical integral component of the Government of Zambia’s tax policy is tax expenditure, intended to ease the tax burden on particular groups of taxpayers and to encourage investment in target areas and activities. However, studies reveal that such policies may lead to tax base erosion, resource misallocation, socially unproductive rent-seeking, opportunities for corruption tendencies, and higher marginal tax rates for compliant taxpayers.

It is therefore imperative to design a framework for estimating and reporting on tax expenditures as part of the annual budget process.

1.1 Definition of Tax Expenditure

Zambia does not have a legal or official definition of tax expenditure. The definition is subjective and differs among tax administrations. According to the UK National Audit Office - TARC (2014), a tax expenditure typically arises when the entitlement to the reduction in tax liability is conditional on the taxable entity possessing specified characteristics or choosing to undertake or forgo a specified action.

For purposes of this study, a tax expenditure refers to potential revenue foregone by the Government as a result of tax relief granted.

Tax expenditures may take the form of:

i. Reductions in the tax rate that allow for application of rates lower than the standard ones;

ii. Reductions to the tax base through exemptions, deductions and allowances; and

iii. Reductions in the tax due through tax credits and waivers. Waivers on fines, penalties and interest do not form part of tax expenditures.

As a general principal, tax reliefs that are structural are not tax expenditures.

Zambia offers a broad range of tax expenditures:

i. To cater for the lower-income sections of the population;

ii. For privileged classes of persons;

iii. For public benefit organizations;

iv. As mandated under regional trade agreements; and

v. To foster investment in priority sectors, activities and locations, under the ZDA Act. These are offered to eligible investors with a minimum investment of US $500,000 and are valid for five years or in perpetuity. The Minister may prescribe additional tax incentives for investments of at least US $10 million in new assets.

Tax expenditures may be statutory or discretionary. The former are granted under the Act of Parliament and are provided for in the principle Zambia tax laws: Customs and Excise Act (First Schedule), Income Tax Act (Second, Fifth, Sixth and Eighth Schedules), VAT Act (First and Second Schedules) and relevant statutory instruments. These principal laws also provide the basis for the standard tax regimes. The tax relief is approved by the Minister of Finance and administered by the Commissioner-General of ZRA.

1.2 Justification for Reporting Tax Expenditures

Tax expenditures are a real financial cost to Government but their obscure nature often leads to little scrutiny and monitoring.

Reporting on tax expenditures periodically has several advantages:

- Gives a holistic view of government spending;

- Informs policymaking and helps identify and estimate the cost of each provision in the tax structure;

- Allows for the evaluation of the efficiency and effectiveness of tax programs; and

- Fosters transparency and accountability.

This report seeks to inform policy and decision-makers of the full cost and realized benefits of the tax expenditures provided in Zambia. It covers tax expenditures related to the four main taxes:

import duty, import excise duty, value added tax and corporate income tax. It excludes personal income tax, mineral royalty tax and fees levied on imports. Due to data limitation, the analysis covers only two years, 2015 and 2016.

2. METHODOLOGY

Three main approaches are used to estimate tax expenditures:

i. Revenue Foregone Method: This is the most frequently applied approach. It estimates the potential tax revenues in the absence of tax expenditures and assumes that taxpayer behaviour remains unchanged when tax policy does. Under the assumption, a business entity would continue to operate normally even when it loses its current tax benefits.

ii. Revenue Gain Method: It estimates the additional revenue when the tax relief is removed, taking into consideration the change in taxpayer behaviour. This approach may be relevant when making estimates of new tax policy proposals for the national budget.

iii. Outlay Equivalent Approach: This estimates direct government spending necessary to substitute for tax relief and give the same benefit to taxpayers as the relief. Like the Revenue Foregone method, it assumes no change in taxpayer behaviour.

The study uses the revenue-foregone method. Although data intensive, this approach is preferred because of its simplicity - assumes no behavioural change - and consistency in the interpretation of results.

2.1 The Benchmark Tax System

This is the basis for defining and estimating the associated tax expenditures. There is no internationally agreed benchmark tax system, although there are some international standards. The benchmark establishes the unit of taxation, tax base and rate structure for the tax period. For each tax type, deviation from the benchmark constitutes a tax expenditure.

2.1.1 Corporate Income Tax

The standard tax rate of 35 percent is considered the benchmark, and every preferential rate to represent a tax expenditure. The Income Tax Act1 spells out the sources of taxable income in a charge year and defines the unit of taxation as an entity that conducts business in order to earn profit.

This analysis assumes that depreciation of an asset aligns with the asset’s useful operating life, and so accelerated depreciation is part of the benchmark.

2.1.1 Customs and Import Excise Duties

There is no hard and fast rule on the benchmark rate. The standard import duty rate of 25 percent is treated as the benchmark.

In the case of excise duty on imports, different rates form part of the normal tax structure depending on the type or category of product per Second Schedule of the Customs and Excise Act. The benchmark was possible where excise rates had a common basis.

1 Under Part III Section 17

Typically, exemptions arising from international treaties or negotiations, such as international organizations and diplomatic agencies, are not treated as tax expenditures. Therefore, preferential rates on imports originating from countries with which Zambia has trade agreements (SADC, COMESA, FTA and SSA) are not considered tax expenditures.

2.1.1 Value Added Tax

As a benchmark, the normal tax structure consists of two rates:

i. Zero percent for exports, based on the destination principle that VAT applies in the jurisdiction of final consumption.

ii. Standard VAT rate of 16 percent for all other supplies.

Sales at minimum taxable value also form part of the benchmark.

Exemption of entities below the statutory threshold for VAT registration is intended to reduce administration costs. It is a component of the benchmark and does not constitute a tax expenditure.

Administration costs for turnover above the threshold are assumed marginal.

2.2 Data Sources

The data used for in the analysis came from the following sources:

i. Firm-level data obtained from Customs declarations (ASYCUDA) and taxpayer return filings to ZRA. The compiled data were then validated by ZRA.

ii. Supply and Use tables (SUTs), which are created by the CSO. SUTs provide data on the flow of commodities and products within the economy, thus allowing for modelling of VAT expenditures across all sectors. A critical assumption was that commodity mix and ratios in GDP remained constant between 2010 and 2016.

iii. Field interviews conducted with ZDA, ZRA, ZACCI and 12 recipients of tax incentives (Annex 9).

2.3 Limitations

Estimation of the tax expenditures had some limitations and caveats:

a) The study relied on the latest data available, so any changes to the underlying data and assumptions could significantly alter the estimates.

b) The assumption that tax policy does not cause changes in taxpayer behaviour may not hold in reality. Therefore, removal of a tax benefit would not necessarily result in a proportionate increase in tax revenue.

c) The effect of each tax type is measured independent of other tax provisions. However, due to interactions among taxes, the total expenditures of any two tax types when treated in isolation could be different from the estimates of the two when combined. For instance, VAT is calculated on CIF and import duty. Thus a preferential duty rate reduces the VAT base and lowers the VAT expenditure. Consequently, summing the tax expenditures of both import duty and VAT would be less than the actual total if the two taxes were treated independently.

d) The study omits negative tax expenditures, which arise when Government collects higher than “normal” revenue; say, when rates higher than the benchmark are applied to the tax base.

Negative tax expenditures would need to be analysed separately from the positive tax expenditures since adding the two could net out the expenditures and create misleading results.

e) The tax expenditure estimates do not take into account non-compliance or tax evasion.

f) Data obtained from ZRA revealed some errors in the classification and description of sectors and products. One example is a company that received approximately 10 percent of the import and excise duty expenditures. It deals in petroleum products but was classified under “Administrative and Support Service activities” sector.

g) Differences in methodology renders international comparison difficult. It would therefore be inappropriate to compare metrics such as tax expenditure to GDP among countries.

h) The sample size of stakeholders interviewed was small and may not be fully representative of the population.

2.4 Corporate Income Tax (CIT)

All business entities are required by law to file income tax returns with ZRA using Schedule 1 (entity’s summary of comprehensive income per source), Schedule 2 (summary of capital allowances), Section A (balance sheet) and Section B (company’s comprehensive income from all sources). First, Schedule 1 was reconciled with Section A and Section B. Schedule 2 was then matched against data in Parts I to IV in order to verify accuracy of capital allowances. Second, gross tax was re-assessed by adding back allowable deductions and allowances (including donations) to the adjusted profit or loss to obtain new taxable income, the benchmark taxable amount. Third, the benchmark tax rate of 35 percent was applied to obtain benchmark income tax.

Tax expenditure was obtained by subtracting the previously assessed tax amount from benchmark tax. Fourth, pivot tables were used to summarise the results.

2.5 Trade-Related Taxes (Customs and Excise Duties)

Both import and excise duties are calculated on either ad valorem or specific rate basis. Given the size and time to extract the data, the request to ZRA for data needed to be comprehensive and precise. Acronyms were clarified and the data validated by ZRA. The data were reformatted to enable calculation of benchmark taxes and tax expenditures for each importation. Pivot tables were then used to summarize the results.

2.5.1 Import Duty

Customs value forms the basis for calculation of import duty. Duty may be calculated on ad valorem basis, i.e. as a percentage of CIF or as specific, in which case the customs value may be CIF value, weight or volume of the item imported.

• Under ad valorem basis, tax expenditure = (CIF x 25 percent) minus customs duty paid

• For specific rates, tax expenditure = (customs value x applicable tariff rate) minus customs duty paid.

2.5.2 Import Excise Duty

This is computed as a percentage of the sum of customs value and customs duty. In this case,

• Under ad valorem basis, tax expenditure = [applicable tariff rate x (customs value + customs duty)] - excise tax paid.

• On specific duty basis, tax expenditure = (applicable tariff rate x excise tax base) - excise tax paid.

2.6 Value Added Tax

VAT is charged at each stage of production or distribution of supplies. VAT expenditures principally result from:

i. exempt and zero-rated imports by final consumers not registered for VAT,

ii. exempt imports by taxable businesses used to manufacture exempt goods,

iii. zero-rated purchases by non-businesses, and

iv. exempt local purchases.

Calculation of tax expenditures on value added is quite unorthodox and tax administrations use different methods. One method that is growing in popularity estimates the VAT expenditure as the difference between the potential VAT under statutory rates and the reference rates. The potential VAT revenue is the sum of the product of potential VAT bases and the corresponding tax rates.

Some of the data in SUTs for use in the estimation model were unrealistic. For instance, the model would yield negative potential value added for the mining sector, which could have resulted from misallocation of mineral exports in compiling the SUT data. In addition, the proportions of output from registered entities did not reflect the structure of the Zambia economy. Under the circumstances, an alternative approach has been used pending revalidation of the SUT data2.

The approach relied solely on taxpayer VAT returns under the assumption that:

a. businesses above the VAT turnover threshold, including those in exempt sectors, deal in standard and zero-rated supplies and file VAT returns in order claim input tax credits; and

b. inter-industry purchases are netted out.

2.6.1 VAT Expenditures on Imports

a. Exempt and zero-rated goods imported by taxable businesses for taxable supplies have no net effect on government revenue because businesses would claim full input tax credit on such supplies.

b. Businesses that deal exclusively in exempt supplies are not eligible to register for VAT in Zambia, so their imports are treated similar to those for final consumers. Such businesses cannot claim input tax credits.

c. Entities making only zero-rated supplies may not register for VAT if approved by the ZRA Commissioner-General. These enterprises would be able to claim full input tax credit on their imports hence no associated tax expenditure. However, since they are not registered, they cannot be identified from others; this may overstate the tax expenditure.

The base for calculation of VAT on imports comprises the Customs value, import duty and excise tax form. Customs value, import duty and excise tax form the import VAT base. Given the availability of import data and that a benchmark rate of 25 percent set as the Customs duty, calculation of import VAT expenditure was based on Customs data.

VAT expenditure on imports was calculated as the sum of expenditure on: (i) imports by non-VAT registered entities, excluding those under trade agreements and diplomatic arrangements and (ii)

2 It is expected that SUT data would be revalidated during the VAT Gap Analysis.

imports by VAT registrants for exempt goods under the trade sector, with a mark-up. Expenditure on exempt imports by registered entities in other sectors is assumed captured under the domestic component.

2.6.2 VAT Expenditure on Domestic Supplies

For domestic production, tax expenditure = [Σz(Yi z - ITCi

z) x τ] + [Σe(Yi e – ITCi

e) x τ]

Yi z = Zero-rated output for activity i

Yi e = Exempt local sales for activity i

ITCi z = input tax credits on output for activity i

ITCi e = input tax credits on exempts for activity i τ = Standard tax rate Exports are excluded.

3. TAX EXPENDITURE ESTIMATES

Estimates reveal that Zambia incurred tax expenditures amounting to K 34.5 billion in 2016 and K 25.7 billion in 2015, representing 16 percent and 14 percent of GDP, respectively.

Contextually, if tax expenditure and direct public spending were computed on the same basis, the amount foregone in 2016 would be equivalent to government expenditure on general public services, economic affairs function and defence that year.

Table 1: Summary of Tax Expenditures

Tax Expenditures 2015 2016

K Mil % of GDP K Mil % of GDP

CIT 9,801.45 5.3% 10,131.07 4.7%

Import VAT 1,294.37 0.7% 1,742.52 0.8%

Import Duty 10,341.77 5.6% 12,546.17 5.8%

Import Excise Duty 114.33 0.1% 4,369.11 2.0%

Domestic VAT 4,194.95 2.3% 5,795.43 2.7%

Total 25,746.88 14.0% 34,584.30 16.0%

GDP (Mil) 183,381.06 216,098.08

Total Tax Revenue 29,435.57 16.1% 29,523.88 13.7%

TE % of Total Tax Revenue 87% 117% Source: ZRA, CSO and MoF.

Tax expenditures are broken down by tax type as below.

Figure 1: Tax Expenditures by Tax Type

10,131.07

12,546.17

4,369.11

1,742.52

5,795.43

9,801.45

10,341.77

114.33

1,294.37

4,194.95

- 2,000 4,000 6,000 8,000 10,000 12,000 14,000

CIT

Import Duty

Import Excise Duty

Import VAT

Domestic VAT

Tax Expenditure (K Mil)

Ta x

Ty pe

2015 2016

Figure 2 gives the distribution by sector, excluding five sectors3 that received less than K 10 million in tax expenditures in each year.

Figure 2: Tax Expenditures by Sector

Notwithstanding differences in methodology, estimates reveal that Zambia has high tax expenditures relative to other Sub-Saharan countries. Figure 3 depicts a comparison of tax expenditure as a percentage of GDP among six countries.

Figure 3: Cross-Country Comparison of Tax Expenditures

Source: World Bank, 2017; *ZRA Data, average for 2015 and 2016 Note: There is variation in the years reported for the countries.

3 Education, Water & Sewerage, Health & Social Work, Entertainment, Household Work

2,000

4,000

6,000

8,000

10,000

Ta x

Ex pe nd itu re (K

M il)

Sector

2016 2015

5.2

1.1

4.0 3.7 4.0

15.0

Ghana Mauritius Rwanda Senegal South Africa Zambia*

TE

a s % o f G

DP

3.1 CIT Expenditures

The main tax expenditures on corporate income are:

- Reduced tax rate on profits from farming, agro-processing, production of fertilisers, export of non-traditional exports and mining operations.

- Deductions for profit on disposal of assets, capital allowances, mining prospecting expenses, mining expenditure, and unrealized foreign exchange gains. Capital allowances consist of capital expenditure on implements, plant and machinery; industrial buildings and commercial buildings; farm improvements, development, farm works and dwelling; and motor vehicles4.

Over 70 percent of the businesses had an effective CIT rate of 0 percent and only 9 percent of them paid at the standard rate of 35 percent in the two years (Table 2).

Table 2: Effective CIT Rate

Effective Tax Rate Proportion of Businesses

2016 2015 0% 73% 71%

1%-15% 11% 12% 16%-30% 5% 6% 31%-34% 2% 2%

35% 9% 9% Source: ZRA Taxpayer returns.

Note: Effective CIT rate is calculated by dividing taxes paid by the actual income. Actual income is the sum of taxable income, exemptions, and deductions.

CIT expenditures totalled K 10.1 billion in 2016, up from K 9.8 billion in 2015, representing 4.7 percent and 5.3 percent of GDP in the respective years.

In both years, three sectors - mining and quarrying, manufacturing, and trade5 - accounted for approximately 60 percent of the expenditures. The mining and quarrying sector was by far the greatest recipient of CIT expenditures, accounting for 37 percent and 30 percent in 2016 and 2015, respectively. Whereas the power sector received 10 percent of the expenditures in 2015, it registered only 0.02 percent in 2016.

The majority of CIT expenditures arose from capital allowances mainly to the mining, manufacturing and agriculture sectors.

Table 3: Main CIT Expenditures (K Mil)

2016 2015 Capital Allowances 5,669.19 6,981.95 Unrealized Forex Gains 1,189.44 1,445.48 Mining Expenditure 507.01 732.15 Losses carried forward 439.46 253.94 Profit on Disposal Assets 66.16 166.73

4 ZRA online system fails to capture capital allowances on motor vehicles, hence probable underestimation of income tax foregone.

5 Trade sector comprises wholesale and retail trade, including repair of motor vehicles and motorcycles.

Figure 4 and Annex 2 show a sectoral summary of CIT expenditures.

Figure 4: CIT Expenditures by Sector

Table 4: Capital Allowances by Sector

2016 2015 Sector Sector

Mining & quarrying 50% Mining & quarrying 67% Manufacturing 10% Trade 6% Finance & Insurance 10% Manufacturing 5% Trade 8% Power 5% Agric., Forestry & Fishing 4% Agric., Forestry & Fishing 4% ICT 3% Transport & Storage 3% Transport & Storage 3% ICT 2% Others 10% Others 8%

Despite receiving substantial tax relief, many firms in these sectors incurred losses as exhibited in table 3. Table 5 indicates the sectoral proportions of tax expenditures in form of losses carried forward.

Table 5: CIT Loss Carried Forward by Sector

2016 2015 Sector Sector

Manufacturing 31% Manufacturing 44% Agric., Forestry & Fishing 18% Mining & quarrying 11% Trade 12% Finance & Insurance 10% Finance & Insurance 9% Trade 10% ICT 9% Agric., Forestry & Fishing 8%

Construction 7% Transport & Storage 3% Mining & quarrying 3% Admin. & Support 3% Others 11% Others 11%

1,000

1,500

2,000

2,500

3,000

3,500

4,000

TE

(K

' M il)

3.2 Trade-Related Tax Expenditures

The tariff structure has 6508 tariff lines at HS 8-digit level and the imports fall under three categories of ad valorem tax rate brackets:

- 5 percent for primary inputs,

- 15 percent for intermediate goods, and

- 25 percent for finished products.

The rationale for this rate structure is that goods expected to have higher local value added should attract less Customs duty in order to encourage local production. This criterion is not adhered to since many finished products also qualify as primary input or intermediate goods.

The data reveals the following bands used: 0 percent, 5 percent, 10 percent, 15 percent, 25 percent, 30 percent and 40 percent computed on CIF; and 11 additional rates6 based on weight or capacity.

Some categories of goods attract specific duty rates of K 2,000 and K 6,000.

3.2.1 Volume of Imports

Zambia’s total volume of imports as measured by CIF value totalled K 72.4 billion in 2016 and K

60.5 billion in 2015.

In each year, 81 percent of the imports attracted no import duty and 12 percent was taxed between 5 and 15 percent. The standard rate of 25% applied to only 5.6 percent of the imports in 2016 and

7.0 % in 2015. Less than 0.1 percent attracted specific rates.

Figure 5: Import Duty Applied on Imports

3.2.2 Import and Excise Duty Expenditure Estimates

In 2016, approximately K 16.9 billion was foregone in import duty and excise duty, compared to K 10.4 billion in 2015. This represented 7.8 percent and 5.7 percent of GDP in the respective years.

6 (0.65, 0.85, 1.0, 1.1, 1.2, 2.0, 2.2, 2.5, 3.0, 3.5 and 4.0)

81.0%

12.0% 5.6%

81.0%

12.0% 7.0%

0%

20%

40%

60%

80%

100%

0% 5-15% 25%

Vo lu m e of Im po rt s, C

IF

Import Duty Applied

Three quarters of these tax expenditures went to four sectors: trade, mining, manufacturing, and administrative services (Annexes 3a and 3b). The trade sector, the largest recipient, accounted for 41 percent and 26 percent of the tax expenditures in 2016 and 2015, respectively.

A large proportion of duty expenditure was on petroleum and bituminous mineral oils, machinery and mechanical appliances, and metal ores and concentrates (Table 5). At product level, most of the duty relief went to petroleum and bituminous mineral oils, approximating 45 percent in 2016.

Motor spirit by itself represented 97 percent of the excise duty expenditures in 2016.

Table 5: Top Products that Contributed to Import and Excise Duty Expenditures

2016 2015

Product K (Mil) % of TE Product K (Mil) % of TE

Mineral fuels, oils and distilled products* 7,603.60 44.70% Machinery and mechanical appliances/parts 1,962 18.80%

Machinery and mechanical appliances/parts 2,351.33 13.80% Mineral fuels, oils and distilled products** 1,786 17.10%

Copper and Cobalt ores and concentrates 1,795.91 10.60% Metal ores and concentrates 1,091 10.40%

Fertilisers 760.19 4.50% Fertilisers 886 8.50%

Pharmaceuticals 707.13 4.20% Vehicles 470 4.50%

Electrical energy 648.1 3.80% Miscellaneous chemical products 385 3.70%

Vehicles 606.28 3.60% Articles of iron and steel 384 3.50%

* In 2016, K 5,571.02 million was from refined petroleum and bituminous mineral oils.

** In 2015, K 1,264 million was from crude petroleum and bituminous mineral oils.

Duty relief from Zambia’s regional trade arrangements amounted to K 6,518 million in 2016 and K 3,455 million in 2015, but these are not considered as tax expenditures.

Table 6: Revenue Foregone from Regional Trade Agreements (K Mil)

Tariff Preference 2016 2015

Import Duty Excise Duty Import Duty Excise Duty

COMESA 0.44 0.01 0.21 0.01

FTA 260.55 2,633.81 320.81 24.62

SDC 380.78 18.53 556.26 7.15

SSA 2,597.93 625.81 2,457.68 87.80

Total 3,239.70 3,278.17 3,334.95 119.57

3.3 Value Added Tax

Zambia’s VAT structure has a broad range of exempt and zero-rated supplies as shown in the appendix. Table 7 highlights a comparison of the country’s structure with those of OECD and the rest of Africa

Table 7: Prevalence of VAT Exemptions and Zero-Ratings in Zambia, Africa and OECD

Percentage of Countries in the region providing Exemptions

Percentage of Countries in the region providing Zero-Rating

Africa OECD Zambia Africa OECD Zambia

Agricultural Inputs 50% 9% X 32% 15% √

Agricultural Produce 71% 6% √ 50% 27% √

Transport 68% 24% √ 39% 30% x

Real Estate 71% 100% √ 7% 3% X

Education 93% 12% √ 18% 0% √

Health/Pharma 79% 94% √ 29% 27% √

Capital Goods 11% 0% √ 4% 9% √

Fuel 7% 0% √ 0% 3% X

Cultural 64% 97% X 4% 27% X

Finance (excludes insurance) 82% 100% √ 7% 6% X

Mining / Petroleum 36% 0% X 11% 0% x

Charitable 7% 97% 4% 6% X

Construction 7% 12% √ 0% 6% √

Tourism inputs/services 18% 6% √ 4% 9% X Source: James. S. and F. Oppong. 2016. Zambia Ministry of Finance.

3.3.1 Import VAT

Import VAT expenditures rose from K 1.29 billion in 2015 to K 1.74 billion in 2016, amounting to 2.3 percent and about 2.7 percent of GDP in the respective years. This includes tax expenditure on public administration and defence of K 42 million in 2015 and 487 million in 2016.

3.3.1.1 Distribution by Sector

In 2016, approximately 82 percent of the expenditures were from four sectors7: Public administration, accommodation and food services; trade; and agriculture. In 2015, 44 percent of the import VAT expenditure was from the trade sector and 22 percent from agriculture (Annex 6).

The products with the highest VAT expenditure on imports are as below.

Table 8: Top Products that Contributed to Import VAT Expenditures

2016 2015

Product K (Mil) % of TE Product K (Mil) % of TE

Pharmaceutical Products 465.87 27% Fertilisers 494.52 39%

Machinery, mechanical appliances & parts* 445.46 26% Pharmaceutical Products 210.02 16%

Fertilisers 375.51 22% Miscellaneous chemical products 174.70 14%

Miscellaneous chemical products 108.50 6% Vehicles 50.30 4%

Vehicles 85.54 5% Machinery, mechanical appliances & parts** 50.22 4%

Railways, tramway locomotives, and parts 29.43 2% Railways, tramway locomotives, and parts 40.23 3% *K 390 mil was from milling machinery in 2016. **Mainly pumps, milling machinery and computers in 2015.

7 Excluding public administration, defence and compulsory social security, the remaining three account for 75 % of the import VAT expenditure.

3.3.1 Domestic VAT

Domestic VAT expenditure was K 5.8 billion in 2016 up from K 4.2 billion in 2015, representing

2.7 percent and 2.3 percent of GDP, respectively. The four largest sectors received over 80 percent of the expenditures in the two years (Table 9).

Table 9: VAT Expenditure on Domestic Supplies

Sector 2016 2015

TE (K Mil) % of TE TE (K Mil) % of

TE

Finance & Insurance 1,286.78 22.2% 955.62 22.8%

Trade 1,216.22 21.0% 909.60 21.7%

Agriculture 970.22 16.7% 838.40 20.0%

Manufacturing 745.18 12.9% 638.13 15.2%

Admin. & Support 587.88 10.1% 472.23 11.3%

Construction 290.82 5.0% 65.47 1.6%

Professional & Science 241.74 4.2% 86.10 2.1%

Transport & Storage 171.35 3.0% 118.76 2.8%

Others 285.24 4.90% 110.65 2.6%

Total 5,795.43 2,945.34

3.4 Tax Expenditure in Relation to Employment

Based on the available employment data8, half of the 20 sectors lost jobs in 2012 and 2014.

Assuming the same average annual change in employment for the two subsequent years (2015 and 2016), Zambia spent approximately K 36,500 in 2015 and K 50,500 in 2016 for every job created.

The largest expense was in the mining and quarrying sector where the amount foregone was K

0.50 million in 2015 and K 0.61 million in 2016 per job lost (Annex 5).

Among the sectors that witnessed increase in employment, financial services was the most costly to government as revenue foregone amounted to K 0.25 million in 2015 and K 0.35 million in 2016 per job created. The most efficient sectors were household activities and education, where approximately K 0 and K 5, 000 was spent per job created, respectively.

3.5 Tax Expenditure in Relation to Profitability

On average, 32 percent of the businesses reported profits during the years (Figure 6). Financial services, manufacturing and real estate had the largest percentage of businesses that recorded profits.

It would be necessary to delve into profitability of the firms and sectors using ratios such as return on capital employed and return on assets.

8 Includes both formal and informal employment.

Figure 6: Firms that Reported Profits

49%

46%

45%

38%

0%

36%

37%

44%

25%

35%

31%

50%

31%

27%

49%

23%

14%

32%

0%

46%

40%

33%

33%

32%

0% 5% 10% 15% 20% 25% 30% 35% 40% 45% 50%

Accomm. & Food

Household Work

Admin. & Support

Agriculture

Entertainment

Construction

Education

Power

Finance & Insurance

Health & Social Work

ICT

Manufacturing

Mining & quarrying

Other services

Professional & Science

Public Admin.

Real Estate

Transport & Storage

Water & Sewerage

Trade

Average

Percentage of Firms Reporting Profits

4. CONCLUSIONS AND POLICY RECOMMENDATIONS

The findings highlight the necessity for Zambia to rationalize its tax expenditure policy. The tax structure appears principally designed to address both systematic and firm-specific issues, rendering it too generous and redundant in several respects.

4.1 Long-term investors consider many of the tax exemptions a mere gift or “add-on”.

Import duty concession is a one-off factor not considered during long-term business planning.

Investors, especially first-movers, need just enough incentivisation to kickstart investment in industries or sectors where hurdles exist.

4.2 Tax expenditures are largely skewed in favour of a few sectors, specifically mining and quarrying, manufacturing, and trade:

a) Whereas mining and quarrying is a highly capital-intensive sector, most of the capital costs are incurred pre-production so many of the tax incentives at the stage of resource extraction are redundant. Smriti Chand (2016) states that once resources are confirmed present, the tax structure is among the least considerations that determine exploitation9.

b) There is little justification for abundant tax relief to the wholesale and retail trade sector

4.3 The current VAT structure and import duty relief on food and agriculture favour higher income households. A 2012 amendment to the VAT Act made some food items normally purchased by lower-income households VAT exempt while wheat and its products, disproportionately purchased by higher-income households, became zero-rated. Since suppliers cannot claim input tax credit on exempt goods, in effect, the change raises prices of the exempts relative to the zero-rated foods, thus making the poor worse off.

Recommendation: Consider zero-rating VAT on cereals grown and processed in Zambia.

Such subsidy would lower consumer prices, widen the distributional reach of stable food to the target population, and incentivize local producers. For basic food imports, it would be more cost effective to replace tax exemptions with direct cash transfers to the vulnerable through, say, food vouchers.

4.4 Administrative loopholes exist in the tax structure making it vulnerable to abuse. Existing firms can access new tax benefits as the old ones expire by forming related businesses under different trading names. In essence, operating costs are transferred from one entity to reduce taxable profits in another.

Recommendation: Limit the ability of companies with subsidiaries to apply tax benefits from one entity to another that does not qualify. One practical way is to separate such entities by type of operation for tax purposes, that is, through ring-fencing.

4.5 The broad range of tax expenditures raises administration costs and constrains ZRA’s resources. This breeds apathy for monitoring “tax-favored” investments, which may explain the absence of comprehensive audit of tax expenditures. Furthermore, the process of phasing

9 The major factors are richness or grade of the ore, size of deposit, method of mining, accessibility, transportation facilities, stage of industrial development, and technology.

out tax holidays poses methodological challenges in apportioning taxable income for investments that have a combination of phased out and existing tax benefits.

Recipients of tax incentives have no obligation to submit returns to ZDA and none reported more than two ZDA visits. The VAT Act permits entities dealing in only exempt supplies not to submit returns to ZRA, while those dealing in only zero-rated supplies can apply for waiver.

Recommendation: Create a rigorous regime of monitoring and enforcement to compel compliance among recipients of tax benefits. Given the volume of tax incentives and the revenue at stake, (i) ZRA should assign a desk to handle tax expenditures; (ii) all recipients of tax incentives be mandated to file accurate and timely returns to ZRA; and (iii) insert provisions in the law for recovery of taxes if tax benefits are abused.

4.6 Tax incentives tend to favor brownfield and foreign-owned investments over greenfield and locally-owned investments, leading to distortions in the economy. Local investors expressed difficulty in raising capital to meet the minimum threshold for tax incentives, a hurdle too short for foreign investors. In the same vein, greenfields have limited cash flows, a key statistic that commercial banks consider, as they tend to lend for working capital.

Additionally, greenfields often fail to utilize income-based incentives since their nascent investments imply longer breakeven period. This leads to unfair competition to smaller firms and greater exposure of the economy to foreign financing.

Recommendation: Government should devise instruments to enable low-capital investors access funding. One way is to replicate and scale a model designed by one foreign investor (Aller Aqua) that partners with commercial banks to introduce outgrowers to capital and knowledge in order to stimulate local supply and competition.

4.7 Copper and cobalt ores and concentrates account for a sizeable proportion of import duty tax expenditures (approximately 10 percent, Table 5). These products are eventually exported with no substantial value added.

Recommendation: Consider imposing standard duty rates on basic metals at importation.

4.8 There is no discernable correlation between tax expenditure and employment or profitability. Sectors that received less tax expenditures performed well in terms of job creation relative to those that with higher tax benefits. Also, under the current tax expenditure policy, about 32 percent of businesses report profits but the distribution varies among sectors.

Recommendation: Tax expenditures should be routinely evaluated for effectiveness and the priority sector list reviewed in line with sector performance. Government should re-examine its policy of shoring up loss-making sectors. At the same time, Government needs to refrain from offering subsidies to businesses that are realising profits.

Recommendation: Tax expenditures should be tailored to creating linkages and positive externalities to stimulate supply and competition.

Recommendation: Consider direct subsidies to firms for labour skill development as alternatives to blanket tax exemptions, especially in labour-intensive or skill-constrained sectors such as the ITC. Government subsidy tied to specific and measurable knowledge and skill transfer would be appropriate. Investment in local human capacity and corporate social responsibility is a compelling business model makes strong business strategy but is costly to implement due to underdeveloped industry linkages.

Recommendation: Explore duty deferment scheme in lieu of import duty exemptions. Duty deferment eases business cash flows while simultaneously protecting revenue since Government loses only the time value of money.

4.9 Many tax expenditures are either perpetual or open-ended and do not specify the legal measures available for the government to recover lost revenue if a recipient violates the terms of the tax benefits.

Recommendation: Insert sunset and claw-back provisions to the tax expenditures.

4.10 There is ambiguity in description of tax-favoured goods and unexplained rationale for some tax relief.

Whereas duty relief on some items appear intended for manufacturers, many of the items are imported by traders. These include iron and steel, cotton and man-made staple fibres.

Differentiation in duty rates for similar products that are indistinguishable poses revenue risk. For instance, cereal seeds attract 5 percent import duty while other cereals are taxed at 15 percent. The data reveal that cereals classified as seeds are chiefly imported under the trade sector rather than the would-be-expected agriculture sector, implying probable misclassification. Some classifications for exemption are broad. For instance, tractors under 8701includes road tractors for semi-trailers (87012000).

In other cases, items that are exempt from duty, such as those under HS codes 8101 to 8501, are imported by final consumers with the ability to pay.

It not clear why some exemptions are offered. Examples include: (a) plastic tubes, pipes and similar items (HS Code 3917); (b) new pneumatic tyres of rubber of sizes 50cm to 57cm; (c) exemption of cooking appliances (HSC 7321) which can be used for both hotel industry and domestic use.

Recommendations: Import duty brackets for items with tax relief should be narrowed, the qualification criteria clarified, and the nature of tax relief clearly described.

Statutory instruments should state the relevant HS codes.

If tax relief is granted on equipment for facilities such as hotels and hospitals, the equipment should be engraved, printed or marked with the institution logos.

For the agriculture sector, the authorities should describe the items covered, such as seeds, machinery tools and implements. If the objective is, say, to provide power to producers, only generators for industrial use should be exempt.

4.12 Investors reported conflicting and onerous bureaucratic procedures, especially for industries engaged in vertical and horizontal integration. Whereas ZRA is charged with administering the tax incentives, ZDA is interested in driving investment volumes.

Recommendation: Encourage inter-agency collaboration to curtail red tape and competing interests, ease information flow and facilitate tax expenditure reporting. In addition, implement non-fiscal measures such as reducing compliance costs for businesses and expedite the VAT refund process to ease business cash flows.

Recommendation: Government should encourage electronic data exchange among the relevant agencies. Greater integration of Ministry of Finance, ZRA, ZDA and CSO data systems would facilitate monitoring and evaluation of tax expenditures. ZRA, the main custodian of data, should maintain clean, easily accessible databases. Tax expenditure analysis is only possible with comprehensive, timely and accurate data.

4.13 Recommendation: Enlist private sector participation in policy formulation and discourse to promote transparency and certainty. Government’s signaling of policy intentions informs private sector planning. Articulation of policy objectives and legal basis of awarding or repealing tax benefits fosters transparency and accountability.

4.14 Recommendation: Reinforce regional cooperation and seek harmonization of tax incentives to enhance fair competition. Incentives should not be used to counter non-tariff barriers imposed by regional competitors to insulate local industries.

Despite the inadequacies, the tax expenditures remain beneficial in a number of areas. Duty exemptions facilitated importation of assets for capital-intensive investments that often have long break-even periods, thus freeing up investors’ liquidity. The tax relief granted to the power sector10 in 2015 appears to have been well-targeted, as the sector did not register any tax expenditures in following year.

It was found that tax relief supported public benefit organizations (PBOs) and enabled them to scale operations for the greater good. Many of the charities rely on foreign assistance, much of which is in kind. Growth of PBOs also implicitly increases their potential to pay income taxes as several have income-generating arms which are liable for income tax, in addition to taxes on employment income.

Noteworthy, investors reported a favourable political environment in Zambia. However, as studies show, investors prefer alleviation of perceived risks and market failures to most of the tax relief:

• Fair playing field to foster competition; tax benefits instead tilt the business landscape.

• Predictable and consistent tax policy. Investors hate surprises, so maintaining stable CIT rates is critical for longer term planning, particularly for industries are sensitive to shocks.

• Clearly spelt out government policies.

• Transparency in tax use.

• Investment protection agreements.

Zambia ranks 85th out of 190 economies on the 2018 World Bank Doing Business Report. Whereas it ranks 15th on the paying taxes indicator, it does not perform well on other indicators. For example, it is 101st on starting business, 149th on registering property and 155th on getting electricity indicators. Addressing these issues would be of vital for attracting and retaining private investment.

10 Electricity, gas, steam and air conditioning supply sector

SUGGESTED FRAMEWORK FOR TAX EXPENDITURE

REPORTING

I. Background Ideally, TE reporting should be codified in the statute.

State the fiscal period covered.

Reasons for tax expenditure reporting.

II. Methodology Define tax expenditure.

Set the benchmark tax structure.

• Benchmark tax system comprises the unit of taxation, taxation period, tax base and tax rate structure.

• No internationally agreed benchmark for any tax type, although there are some agreed principles.

• Basic principles:

- Similar taxpayers or activities are subjected to standard treatment; tax treatment should not favour or disadvantage similar taxpayers or activities.

- Benchmark may include structural elements of the tax system, such as progressive income tax rates, may be included in the benchmark.

Prepare detailed inventory of tax expenditures, including those covered in the law, cabinet directives and statutory instruments.

Map out current and previous procedures for claiming the tax benefits.

Indicate limitations to reporting.

III. Summary of Tax Expenditures Estimates Break down tax expenditure by tax type

- Outline the tax provision granting the tax expenditure, description of the relief and nature or category of items, revenue forgone, express as percentage of revenue foregone (percentage of tax collected), and as percentage of GDP;

- objective of the relief; date enacted or effected;

- sunset clauses;

- schedule for review;

- metrics for assessing effectiveness, e.g. number of jobs created.

Show trends in TE estimates, by function and by size (in measured TEs.

Reveal tax expenditures that benefit and those that penalize sections of taxpayers.

Show large negatives.

List beneficiaries

- Entities and sectors by amount (may indicate, say, top 20 percent).

Ideally, only tax expenditure with reasonable margins of error should be considered.

Tax expenditures include waivers and reimbursements but not waivers on penalties and interest.

IV. Dissemination Report should be readily accessible in print and online.

For effectiveness, the report should be ready (prepared in time) for annual budget and policymaking.

ANNEXES

Annex 1: Volume of Imports by Import Duty Rate (2015-2016)

Import Duty Rate 2016 2015

CIF Kwacha % of CIF CIF Kwacha % of CIF 0 58,670,384,669 81.1% 49,021,729,037 81.0%

0.65 72,685 0.0% 40,405 0.0%

0.85 380 0.0% 913 0.0%

1 677 0.0% 2,280 0.0%

1.1 29,649,516 0.0% 31,520,059 0.1%

1.2 571,927 0.0% 247,015 0.0% 2 134,046 0.0% 670,477 0.0%

2.2 - 5,987,878 0.0%

2.5 2,802,656 0.0% 5,298,905 0.0% 3 9,154,081 0.0% 6,259,624 0.0%

3.5 42,126 0.0% 355,697 0.0% 4 1,246,658 0.0% - 5 3,749,957,342 5.2% 2,791,510,176 4.6%

10 11,314,739 0.0% 2,705,589 0.0% 15 4,631,840,497 6.4% 4,515,987,946 7.5% 25 4,145,550,146 5.7% 4,116,348,235 6.8% 30 909,902,702 1.3% 26,308,425 0.0% 40 77,622,790 0.1% -

2000* 8,398 0.0% 314,686 0.0% 6000* 25,009,020 0.0% -

Not stated 85,391,848 0.1% - Total CIF 72,350,656,903 60,525,287,345

* Refers to Specific tax rate. Source: ZRA, ASYCUDA

Annex 2: Estimated CIT Expenditures

Sector 2016 2015

TE (K' Mil) % of TE TE (K' Mil) % of TE Mining & quarrying 3,746.90 37.0% 2,925.48 29.8% Manufacturing 1,208.10 11.9% 1,193.28 12.2% Trade 1,109.28 10.9% 1,048.62 10.7% Other services 867.03…