Some competition sharpens everyone it touches: better products, honest prices, faster service, the discipline of craftsmen measuring themselves against craftsmen. And some competition is a fire in a shared building, where every move destroys more total value than the winner ever captures. The name for the second kind is negative-sum, and the tragedy of it is that each player’s moves are individually rational while the game’s outcome ruins them all. The matatu route where competing crews race dangerously for the same passengers; the printing street where quotes have fallen below material cost; the two water companies digging parallel pipe networks neither can fill; the salary bidding war for the region’s five competent accountants. Recognizing negative-sum dynamics early, from inside the fog of daily rivalry, is one of the most valuable diagnostic skills an operator can build, because the only winning move in such games is to change them.
Key Takeaways
- Negative-sum competition destroys more total value than it allocates: the fight’s cost to all players exceeds the prize’s worth. Individually rational moves, collectively ruinous outcome.
- The three classic forms: price wars in trust- or relationship-limited markets, duplicated infrastructure neither rival can fill, and bidding wars for scarce inputs like trained talent.
- The signatures are measurable: prices below unit cost, capacity utilization falling across all rivals as capacity grows, input costs rising faster than output prices, and margins shrinking for every player at once.
- The trap holds because unilateral exit is punished: the firm that stops the price war alone loses share before the market corrects. Escapes are therefore usually collective: standards, floors, shared infrastructure, or changed rules.
- The PARTS levers are the toolkit: change players, raise added value through differentiation, change rules through associations, signal credibly, or change scope by exiting to a different game.
- Price competition itself is not the villain; healthy markets need it. The line is crossed when the mechanism destroys the market’s capacity to serve anyone: quality collapse, safety collapse, trust collapse.
How does rational rivalry produce collective ruin?
Through a structure game theorists know well: each player’s best response to the other’s move worsens the joint outcome, and no one can stop unilaterally. The price war is the canonical case. One printer cuts prices to fill idle machines; rivals must match or watch customers walk; the new price level fills no more machines than before, because total demand for printing did not grow, but it funds no maintenance, no training, and no honest wage. Every shop is now working harder for less, quality decays as corners get cut, and customers, taught by the war that print is a commodity, punish any attempt to restore sane prices. The market has been collectively poorer for years, and every individual decision along the way was defensible.
Duplicated infrastructure runs the same logic in capital. Two rivals each build the warehouse, the fleet, the network that the market’s volume can only fill once. Both operate at half utilization; both carry full fixed costs; both price desperately to chase the volume that would fix their utilization, importing a price war into the capex war. East Africa has seen the pattern in telecom towers before infrastructure sharing, in parallel distribution networks, and in every trading centre where the fourth identical hardware shop opens beside three struggling ones. The talent version is quieter: firms bidding up the same five certified accountants raise everyone’s cost base without adding a single accountant to the region, the textbook negative-sum input auction, whose real solution, training more accountants, is a floor investment no single bidder has the incentive to make alone.
The reason exit is hard is the reason the trap is a trap: unilateral peace is punished. Stop cutting prices alone and you lose share immediately while the benefit, a healthier market, arrives slowly and accrues to everyone including the rivals still fighting. Games with that payoff structure do not end by individual virtue. They end by design.
How do you diagnose it before you are deep inside?
Four measurable signatures, best reviewed quarterly with the operating cadence’s honesty.
Prices below defensible unit cost, sector-wide. Not one desperate discounter, but a market whose going rate cannot fund quality delivery by anyone. When winning a job means losing money carefully, the war is on.
Utilization falling as capacity grows. Count the sector’s idle machines, empty trucks, half-full warehouses. If every rival’s utilization drops while total capacity rises, the market is paying for infrastructure it cannot feed.
Input costs rising faster than output prices. The talent-war signature: salaries, rents, or raw material bids climbing across all players while customer prices stay flat or fall. The margin between the two auctions is being ground out of everyone.
Universal margin compression. In healthy rivalry, someone is winning: share shifts, and the better operator’s margins hold. When every player’s margins shrink together for seasons, the game itself, not any player’s play, is the problem, and the pie is shrinking rather than being divided.
What are the exits?
All of them are game changes, which is why this essay sits in a coopetition series; the PARTS levers are the toolkit, and most require company.
Differentiate out (Added value). The single-firm escape: stop selling what the war is pricing. Reliability, speed, bundled service, a niche segment, anything that makes your offer non-identical moves you out of the auction. It is the right first move and often insufficient alone, because thin markets can drag near-substitutes back into the fight.
Set floors together (Rules). Associations exist for exactly this: minimum quality standards, published rate cards, credit-conduct codes. Done as quality floors and transparency, these are legitimate market repair; done as secret price-fixing, they are illegal and deserve to be. The line is public, quality-linked, and customer-serving versus covert and extractive. The layered-rival ledger discipline applies in full.
Share the infrastructure (Players and Scope). Where the war is duplicated capex, the exit is the shared asset: the common cold store, the tower company, the joint delivery network, rivalry moved up the stack onto service while the floor is co-owned. This is floor-and-ceiling doctrine applied as rescue rather than foresight.
Grow the scarce input (the long game). Talent wars end when the talent pool grows: apprenticeships, sector academies, shared training costs. The firms that fund the pool capture loyalty and first pick, and the war quietly dissolves for everyone.
Leave (Scope). Some games are not worth winning at the price the war sets. Redeploying capital to a market that rewards quality is not defeat; it is the Disappearance Test run on the game itself: what would we lose by leaving this fight, and is it more than we are losing by staying? For the steward, whose capital and people are entrusted rather than owned, refusing to feed a value-destroying fire is not timidity. It is the job.
FAQ
What is negative-sum competition?
Rivalry whose total cost to all players exceeds the value of the prize being contested: price wars below cost, duplicated infrastructure neither rival can fill, and bidding wars for scarce inputs. Each move is individually rational; the collective outcome impoverishes everyone.
How do I know my market is in a race to the bottom?
Four signatures: sector-wide prices below defensible unit cost, falling utilization as capacity grows, input costs rising faster than output prices, and margins compressing for every player at once rather than share shifting to better operators.
Why can’t one firm just stop the price war?
Because unilateral exit is punished: the firm that restores sane prices alone loses share immediately while the benefits of a healthier market accrue slowly to everyone. Escapes are structural or collective: differentiation, standards, shared infrastructure, or exit.
Are industry price floors legal?
Quality standards, transparent rate cards, and conduct codes serving customers are legitimate market repair. Secret price-fixing that extracts from customers is collusion and illegal. The test is public, quality-linked, and customer-serving versus covert and extractive.
When is leaving the market the right move?
When the war’s price of participation exceeds what winning could ever return. Run the Disappearance Test on the game: what do we lose by leaving, and is it less than staying costs? Redeploying entrusted capital from value-destroying fights is stewardship, not defeat.
Related Reading
- The Pie Is the Point: Value Creation vs Value Capture
- Change the Game, Don’t Just Play It: The PARTS Framework
- Co-build the Floor, Compete on the Ceiling
- Pricing: The Neglected Lever of the African SME
Sources and Evidence
- Brandenburger and Nalebuff, “The Right Game: Use Game Theory to Shape Strategy,” Harvard Business Review (1995): game structures where rivalry destroys joint value and the case for changing the game.
- Co-opetition (Brandenburger and Nalebuff, 1996), overview: the win-lose and lose-lose dynamics the framework addresses.
- GSMA, “Understanding mobile money interoperability”: shared infrastructure as the exit from duplicated-capex rivalry.