How East African Tax Wars Are Hurting the Poor

1. A Beggar-Thy-Neighbour welcome mat

From the Indian Ocean coastline of Mombasa to the shores of Lake Victoria, a quiet but destructive contest is under way. Kenya, Tanzania, Uganda and Rwanda are out-bidding each other with ever-sweeter tax holidays, import-duty waivers and zero-rating deals meant to lure foreign investors. The result is a region-wide “race to the bottom” that is draining public coffers of up to US $2.8 billion a year—money that could have financed schools, rural clinics or safety-net programmes for the 80 million East Africans who still live on less than US $3.20 a day

2. What the “Tax Wars” Look Like on the Ground

• Kenya’s Export-Processing Zones (EPZs): 10-year corporate-income-tax holiday, zero import duty on raw materials, no VAT on inputs.
• Tanzania’s Special Economic Zones: Identical 10-year tax holiday plus exemptions from all local-government levies; mining companies enjoy fuel-duty waivers and capital-gains-tax exemptions locked in by “stability agreements” that can run for 25 years.
• Uganda: 10-year income-tax holiday for agro-processors and exporters that sell at least 80 % of output abroad; undisclosed production-sharing agreements with oil majors cap their tax exposure for two decades.
• Rwanda: Foreign investors bringing in ≥ US $250 000 automatically qualify for VAT, customs and withholding-tax exemptions; 84 % of the forgone revenue comes from import-duty relief, only 0.17 % is tied to job-creation targets

The arithmetic is brutal. In Rwanda, incentives worth 4.7 % of GDP in 2009 could have doubled the health budget. Tanzania’s US $1.4 billion annual revenue loss is six times what it spends on agricultural extension services; Kenya’s US $1.1 billion giveaway equals its entire health-sector allocation

3. Who wins? Who loses?

Winners: A narrow cohort of multinational manufacturers, agribusiness conglomerates and oil & mining houses that would probably have invested anyway. IMF surveys show only 1% of firms cite EPZ tax breaks as the principal reason for choosing Kenya; far bigger draws are market access, infrastructure quality and political stability

Losers:

1. Poor households—when VAT is near-uniform at 18 % but essential goods are not zero-rated, the tax burden becomes disproportionately regressive.
2. Micro- and small-traders—left outside the ring-fence of incentives yet still paying presumptive taxes and licence fees.
3. Women and girls—who rely most on under-funded public health and water services.
4. Future generations—because every dollar foregone today is a dollar plus interest not invested in schools, roads or climate resilience.

4. Why the Race Is Accelerating

East African Community (EAC) treaty language calls for “harmonisation” of investment incentives by 2008; the deadline came and went. A Draft Code of Conduct against Harmful Tax Competition has sat on the Council of Ministers’ desk since 2010. National investment agencies therefore operate in a coordination vacuum, using tax breaks as the easiest—if costliest—differentiator. The secrecy surrounding individual mining and oil agreements makes cross-border transparency even harder.

Illicit trade is an extra accelerant. When neighbouring states apply sharply different duty rates on the same product, smugglers arbitrage the gap—eroding the legitimate tax base and encouraging yet more exemptions to “compensate” formal firms.

5. The False Promise of “Crowding-In”

Proponents argue that tax holidays crowd-in foreign capital, create jobs and eventually broaden the tax base. Evidence from the past 15 years says otherwise:

• Tanzania’s EPZ law of 2002 “has not resulted in a noticeable pick-up in FDI”
• Uganda attracts the highest green-field FDI in the region despite offering the fewest incentives.
• World Bank analyses show tax-driven investors are “foot-loose”: garment firms closed shop in Kenya’s EPZs almost immediately after the 10-year holiday expired, shifting to lower-cost neighbours

6. A Better deal for citizens: Four policy shifts
7. Publish every incentive
   Governments should table an annual Tax-Expenditure Report in parliament detailing the cost, duration and beneficiaries of every exemption—oil production-sharing agreements included.
8. Scrap holidays, keep accelerated depreciation
   The IMF, OECD and AfDB agree that accelerated depreciation and loss-carry-forward provisions are less distortionary than outright tax holidays. East Africa should phase out blanket income-tax exemptions and replace them with investment allowances tied to verifiable capital expenditure.

• Agree a regional minimum corporate-tax floor
  The EAC can borrow from the West African Economic and Monetary Union (WAEMU) which enforces a 15 % minimum corporate rate. A 15–20 % band—with no holiday longer than three years—would curb destructive competition while still allowing smaller states such as Rwanda to offer modest location advantages.

1. Tie remaining incentives to social returns
   Make eligibility conditional on measurable development outcomes: number of local jobs created, share of local procurement, or contribution to green technology. Sunset clauses must be non-renewable and subject to public audit.

7. Time to Choose: Region or Ruin?

East Africa stands at a fork in the road. One path leads to deeper integration—a single customs territory, shared infrastructure and coordinated tax policy that protects public revenues. The other path continues the current beggar-thy-neighbour spiral, shrinking fiscal space just when populations are booming and climate shocks demand public investment.

The cost of inaction is already etched in crowded classrooms, over-burdened clinics and potholed highways. Ending the tax war is not an attack on investors; it is a prerequisite for inclusive, resilient growth. As the EAC heads of state prepare their next budget speeches, they should remember: every shilling, franc or shilling surrendered in secret to a foreign boardroom is a shilling stolen from the region’s poorest citizens. The race to the bottom only ends when leaders stop running—and start governing.