The 294,000-Birr Square Metre
In mid-2023, after a five-year break, Addis Ababa put 297 plots of public land up for lease auction. One was a 390 m² plot in the Sar Bet area of Kirkos District, where land usually trades around 90,000 birr per square metre. The top bid was 695,000 birr/m², with 43% paid up front. The plot went to a bid of 401,000 birr/m², paid in full on signing: 156.3 million birr (Addis Fortune). The city turned down 294,000 birr per square metre, about 42% of the top offer. The question is what rule makes that the right answer.
The rule on paper. Ethiopia’s urban land is constitutionally public, and since Proclamation 80/1993 cities have leased it out by tender. The current federal law, Proclamation 721/2011, says the winner is chosen “on the basis of his bid price and the amount of advance payment he offers” (Art. 11(5)). Art. 20 sets a minimum down payment of 10%. The balance is paid “in equal annual installments,” with interest on what remains at the Commercial Bank of Ethiopia’s lending rate. Weldesilassie and Worku’s study of 3,038 Addis plots from 2013 to 2017 (paper) gives the city’s weights: bid price 80% of the score, down payment 20%.
Testing the rule. Normalise each component to the best offer in the field. The 695k bid then scores 80 + 20Ă—0.43 = 88.6, and the 401k cash bid scores 80Ă—(401/695) + 20 = 66.2. So the published rule picks the bidder who lost. It also gets every other contested result in the same reports wrong:
- Arada: a 100,000 birr/m² bid with 50% down beat a 105,000 bid with 40% down. Under 80/20 the higher bid wins.
- Arada again: a 60,100 cash bid beat a 106,000 bid, which means the city accepted a 43% discount.
- Nifas Silk: a 65,180 cash bid beat a rival “who placed double.”
- In 2024, Awash Bank won a plot near Teklehaimanot at 311,000 birr/m² paid up front, beating a record 350,000 bid (Addis Fortune).
One rule fits all five with no exceptions: rank by birr per square metre paid at signing, i.e. price × down-payment share. Sar Bet: 695k × 0.43 = 299k against 401k, so cash wins. Arada: 105k × 0.40 = 42k against 100k × 0.50 = 50k. For the 60,100 cash bid to win, the 106k bid needs under 57% down, and the floor was 40–43%. For Awash, the 350k bid needs under 89% down. I can’t see the bureau’s internal scoring directive. What I can show is that the 80/20 formula fails these published outcomes and an upfront-cash ranking reproduces all of them. Fortune’s reporter in the room called it “the City officials’ preference for immediate liquidity.” By February 2026 bidders had worked this out. The minimum was 40% down, but bidders were “increasingly paying 100pc at auction” (Addis Fortune).
Is cash worth that much? By 2024 the terms were 40% down and the balance within five years. So I priced the 695k bid’s actual payment stream: 43% at signing, then the remaining 57% in five equal principal instalments plus interest on the declining balance. I deflated that by the IMF’s inflation path, the sharpest erosion the city could plausibly face: 29.1% (2023), then 20.7%, 16.5%, 12.6% and about 10%.
| Interest charged on balance | Real PV of the 695k bid | Share of face | Discount rate that equates it to 401k cash |
|---|---|---|---|
| 0% | 529k | 0.76 | 72%/yr |
| 10% | 604k | 0.87 | 107%/yr |
| 16.5% | 653k | 0.94 | 131%/yr |
Take zero interest and Ethiopia’s worst recent inflation, and the instalment bid is still worth about 529k in real terms. That is roughly 128k/m² more than the cash bid, or about 50 million birr on this one plot. For the city to be indifferent between the two, it would have to discount money at 72–131% a year. Meanwhile the Treasury was recently borrowing for four weeks at 1.95% (Addis Fortune editorial). Default risk doesn’t rescue the choice either. The city keeps the down payment, and Art. 20(6) lets it seize the property after three years of arrears. Even so, ranking by cash only breaks even if about 67–71% of instalment winners stop paying. The best figure available comes from the 2024 round, where about 70 of 216 winners (32%) failed to sign by the deadline (Capital). That failure happens before any money changes hands, and the law already handles it by passing the plot to the runner-up.
What it adds up to. In practice the auction ranks bidders by how much cash they hold, not by how much they value the land. The winners show it: Awash Bank, Ethiopian Reinsurance (102,213/m² in cash, ahead of the runner-up by 2,113), Ajewa Real Estate. This fits land scholar Wudu Muluneh’s account of a lease system that keeps rebuilding its legal machinery while value leaks out through administration. Here the leak shows up in the city’s own published results. Auctions are also a small door. The mayor reported 4.83 million m² transferred in a year, of which 23.9 ha, about 5%, went through public auction. That 5% is the only part of the system where the public gets to see a price.
The bet. A 2024 draft proclamation would rescore tenders as 65% price, 20% advance payment and 15% lease payment over the term (Addis Insight). My bet is that published Addis award lists under the new law will still be explained by upfront birr per m² in at least 9 of 10 contested plots. The falsifier is easy to spot: a reported award where a higher-priced bid with less upfront cash beats an all-cash rival. If that starts happening, the city has started pricing its land by value rather than by liquidity.
Sources: Addis Fortune (2023 auction, 2024 auction, 2024 record bid, Feb 2026 round, Oct 2026 editorial); Proclamation 721/2011 (Arts. 11, 20); Weldesilassie & Worku, “Bidders Premium on Government Land Auction”; Capital Ethiopia, June 2024; Addis Insight on the draft proclamation; IMF WEO inflation figures; Wudu Muluneh, Land Economics blog. Caveat: the scoring rule is inferred from five reported contested outcomes, not read from a published directive.