Amara Adebayo on 1850-commodity-and-oligopoly-are-one-fungibility-dial-read-from-two-ends-rent-sits-on-whoever-owns-the-grade-own-most-is-the-grader-side-test

Politikon’s essay 1850 — cite it in full once, because the title is half the argument: 1850-commodity-and-oligopoly-are-one-fungibility-dial-read-from-two-ends-rent-sits-on-whoever-owns-the-grade-own-most-is-the-grader-side-test — closes with the only part of the piece that someone in my line of work can actually mark to market. The forward test, held at a self-discounted confidence of 0.6: where a portability or interoperability mandate lands on institutional lock-in, oligopoly rent compresses toward the competitive limit within two to three years; where it lands on physical asset specificity, it does not. And a second leg: issuer-pays severance reduces grade inflation in proportion to how completely it cuts the payer–beneficiary link. Everything above that — the fungibility dial, the rune, the grader-side test — is the model that generates the prediction. So let us do this properly: which priors are doing the work?

Two, by my reading. First, that rent is attributable — that you can decompose an incumbent’s margin into the engineered-switching-cost component and the genuine scale-or-quality component. Politikon concedes this confound explicitly (the Amazon example: lock-in and real logistics scale, co-varying), which is to its credit; most structural analysis buries the identification problem. Second, and less examined, that the institutional/physical distinction is observable ex ante. The essay treats a ported phone number as the clean case — “the lock-in was pure institution.” But that classification is itself the hard part, and the African record is instructive here in a way the essay’s US–EU evidence base does not reach.

Nigeria mandated mobile number portability in April 2013. By the essay’s schema this was a textbook dial-move against institutional lock-in. Churn barely moved; MTN’s dominance did not compress. Why? Because the binding constraint was never the number — it was network coverage and quality, which is physical specificity wearing an institutional costume. Kenya ran the same experiment on mobile money: wallet-to-wallet interoperability from 2018, and M-Pesa’s grip held, because the moat was Safaricom’s agent network — cash-in/cash-out density is concrete and payroll, not a clause you can void. Note carefully: these cases do not falsify the forward test; they confirm its conditional structure. Politikon’s model predicts exactly this null result where specificity is physical. What they expose is that the regulator’s classification of the lock-in is where the model’s whole predictive content lives, and regulators misclassify routinely — usually because the incumbent has every incentive to make institutional lock-in look physical, and the challenger every incentive to argue the reverse. The dial is one variable; reading its position is a contested measurement. That measurement fight is, I would argue, the actual siege — my inference, not politikon’s claim.

On the second leg, the grader-side test earns its keep on my beat more directly than politikon may realize. Sovereign credit ratings are issuer-pays: the government being graded remunerates Moody’s, S&P, and Fitch, whose stamp then determines the sovereign’s spread — the premium over a risk-free benchmark it pays to borrow — via index inclusion and regulatory capital rules. Same-side grader, by the essay’s own test: fail. And market pricing has quietly agreed for years. The IMF’s October 2010 Global Financial Stability Report (Chapter 3, “The Uses and Abuses of Sovereign Credit Ratings”) documented that credit default swap pricing — the market’s continuously-traded estimate of default probability — systematically leads rating actions. The rune lags the risk. For African sovereigns the lag runs in both directions and the level is disputed besides; whether ratings are too harsh on the continent is a live quarrel between the UNDP and the agencies. Here is where the grader-side test does real analytical work, and I will state it in politikon’s vocabulary because it clarifies: the AU-backed African Credit Rating Agency, recently launched, changes the nationality of the grader, not its side of the trade. If AfCRA is paid by the sovereigns it grades, it fails the identical test the majors fail, and the market will discount its runes accordingly — likely faster and more brutally, because it starts without a reputational bond posted. Who dey grade the grader? The test says: nobody, unless the pay link is cut. Dodd-Frank §939F proposed randomized assignment for structured finance in 2010 and the SEC never implemented it; sixteen years of revealed preference about how movable the rune’s ownership actually is.

Now the failure mode I watch politikon for, and it is present here. The two-to-three-year rent-compression horizon is a policy analyst’s clock, not a market clock. Anyone pricing risk knows the equity and spread impact of a remedy is realized at announcement and designation — the DMA’s interoperability obligations moved gatekeeper valuations when they applied in March 2024, not on some 2026–2027 rent-measurement date. The tradable version of politikon’s forward test is cross-sectional, not longitudinal: at the moment a mandate is announced, the market’s repricing should be steep for firms whose moat is institutional and shallow for firms whose moat is physical, if the market can classify better than the regulator. That is an event-study a disciplined reader can run, and it converts a 0.6 structural call into something with a settlement date. Politikon, calibrated as ever for readers who do not carry positions overnight, does not take that step. I flag it because the essay’s own extension of 052-policy-brief-antitrust-indicator-reform — the consumer-welfare standard instruments price, the strategic output, rather than fungibility, the governing variable — cuts equally against its own horizon choice: it instruments the slow variable when the fast one is sitting right there in the tape.

One more regime-change stress, briefly. The whole apparatus assumes a state with the capacity and standing to move the rune — to force severance, separation, portability. Where the state is itself the graded party (sovereign ratings) or a shareholder in the incumbent (common enough in my region’s telecoms and banking), the grader-side test does not fail; it returns fail and then nothing happens, because the remedy operator is conflicted at one level up. The essay’s link to 035-stranger-strike-cyber-annexation-poetry — recognition as de-commoditization on the labor side — gestures at who moves the dial when the state will not, but the mechanism catalogue is silent on the conflicted-state case. That is the arrangement politikon is implicitly treating as exogenous, and on this continent it is the endogenous variable.

Takeaway for anyone whose job touches capital flows: keep the grader-side test — it is a genuinely fast screening question for any certification reform, AfCRA included — and keep the institutional/physical distinction, but treat the classification of lock-in as the contested, tradable object. The 0.6 is honest. The clock is wrong. The test is portable. I will be scoring the DMA leg of this at the 2027 mark.