Le texte de l’ouvrage est en anglais, tel qu’il a été écrit ; seule l’interface du site est en français.

Annex I — The Tunisian case file: evidentiary status, classification under Annex F and Configuration M trajectory

Annexe 9 sur 17 · ≈ 10 min de lecture · 19 paragraphes · 31ᵉ des 40 sections de lecture · rechercher dans l’ouvrage

Case file, companion to §9.3, to Annex F and to Annex G¶ 1

Status and scope. This annex consolidates in a single place everything the document asserts, and refuses to assert, about the Tunisian case of monetary spillovers. It is the case file provided for by the transitional provision of Annex F (§F.16): it fixes the evidentiary status of each figure, the classification of the case under the Annex F framework at the pre-Configuration M level, the reconciliation obligation that conditions any statement of convergence, and the programme of work leading to the Configuration M computation — the first deliverable of the empirical programme of this report (Annex I; international extension, Annex J). It produces no new estimate, identifies no causal effect, and establishes neither welfare loss, nor legal responsibility, nor price. All the magnitudes below are reported under a single inherited counterfactual, denoted c_V5 (policy rates held at their end-2021 level, of the c₀ family), in nominal Tunisian dinars cumulated over 2022–2025, without rebasing — a declared derogation from §F.8. Every figure in this annex is a programme scenario within the meaning of §F.16.¶ 2

I.1 Why Tunisia, and the author’s position

Tunisia is the recipient-country profile for which the fifteen-mechanism decomposition of Section 9 was specified: sovereign external debt predominantly denominated in foreign currency, a partially open capital account, structural dependence on energy and food imports, and the absence of a fiscal buffer. The case is chosen for the visibility of the broken interface, not for its severity, and Section 9 states the epistemic consequence: were this case to fall, the chain would lose a family, not its foundation.¶ 3

Positional disclosure. The author served as Minister of Finance of the Republic of Tunisia at the outset of the Covid-19 crisis, and the case study concerns that same country. This proximity is a source of field knowledge and a legitimate ground for scrutiny: the reader is entitled to ask whether the choice of case, the counterfactual or the attribution bounds lean in favour of the recipient country. The annex answers by structure rather than by assurance. Every constraint that applies to this case works against inflation of the claim: the conservative zeros of Table F2 (A5, B1–B2 and the supply-origin component 5b), the strictly domestic restriction on A7, the sign convention of C3 — which requires carrying a favourable tourism component alongside the unfavourable one — the refusal of any offsetting between agents without a documented transfer, and the requirement that the joint attribution set be pre-registered. The decisive mitigation is nonetheless external: the computation protocol is designed to be executed and replicated by parties unconnected with the author, and the programme is already committed to a second recipient country, of a different profile, within a regional panel (Annex G), so that no conclusion rests on the author’s country alone.¶ 4

I.2 The evidentiary register

Three sets of figures exist for this case, and this annex is the place where the status of each is fixed, so that no reader can confuse them.¶ 5

The programme workbook (V5). The programme’s internal workbook decomposes the 2022–2025 tightening cycle into channels, applies national-level offsets and a global overlap discount, and arrives at an all-causes gross figure of about 15 billion TND (≈ 4.5 billion euros) and a consolidated national net of 2.98–11.73 billion TND (central value 6.45). This consolidation offsets gains and losses across agents, which the Annex F framework does not permit; the workbook therefore serves as source material for the case, not as a result.¶ 6

The monetary core of §9.3. The body of the document reports a strictly monetary core, net of offsets, of 0.76–1.0 billion euros. Under §F.16, this figure is a programme scenario retained in order to make the anatomy of Section 9 legible; it grounds no claim beyond illustration, and each of its occurrences in the body refers back to the present annex.¶ 7

The classification under Annex F. The case file maps the workbook’s channels onto the Section 9 taxonomy, applies the exclusions and separations of Annex F, and reports the per-agent scenario ranges summarised in Table I1. This is a classification, not yet a computation proper to the framework: the underlying values are inherited from the workbook under the single counterfactual c_V5, and the low/base/high bands are scenario bands, not constrained extrema over a pre-registered joint set.¶ 8

I.3 Mapping the channels under the Section 9 taxonomy

Table I2 summarises the mapping of the case’s channels onto the Section 9 taxonomy and the restrictions of Table F2. Three features of this mapping deserve attention: the essential-imports channel separates into a monetary exchange-rate pass-through component (5a) and an excluded supply-origin component (5b); the private credit channel belongs to the cross-border bank risk-taking family, distinct from the domestic crowding-out channel A7; and C3 is carried as an unfavourable income component and a favourable price component, its net direction being an empirical result.¶ 9

I.4 What the framework requires of this case, and what it finds

Three consequences follow from the application of the Annex F framework to this case, and one substantive finding.¶ 10

Distribution becomes primary. The result is a per-agent vector — Treasury, central bank, firms, households — and any figure summing several agents is labelled accounting consolidation (§F.9). Because undocumented gains no longer offset the losses of other agents (T = 0 by default), the accounting consolidation is mechanically slightly higher than a conventional national net at equal scope; the difference is not a revision of the facts, but the price of refusing to offset a loss by a gain that never reached the party bearing the loss.¶ 11

The exclusions become visible. The supply-origin component of the import bill (5b: 338–1,650 million TND, base 800) is excluded from any attributable aggregate and displayed as such, rather than silently merged; the pre-existing idiosyncratic premium is held at zero; the revaluation of the debt stock remains a memo item.¶ 12

Sensitivity becomes explicit. The private-sector scenario visibly depends on a declared working assumption — the exchange-rate pass-through share of the import bill, set at 0.50 pending documentation by primitive components. Any presentation must show the scenario at alternative shares (0.30 and 0.70 at a minimum) as long as the split remains undocumented.¶ 13

The finding. Applied to this case, the framework does not widen the defensible claim; it narrows it and brackets it. The direct external cash-flow core (base ≈ 0.68 billion euros) lies below the monetary core of §9.3 (0.76–1.0 billion euros), and the extended Treasury incidence (base ≈ 1.04 billion euros) brackets it from above. A framework designed by a former Minister of Finance of the country under study, applied to that country, reduces the headline figure instead of inflating it — which is precisely the behaviour that the credibility of the architecture requires, and the substance of the statement of convergence or divergence that §F.16 obliges the programme to publish once the reconciliation of I.5 is closed.¶ 14

I.5 The reconciliation obligation

The consolidated national net of the programme workbook (central value 6.45 billion TND ≈ 1.93 billion euros) and the monetary core of §9.3 (0.76–1.0 billion euros ≈ 2.55–3.35 billion TND) do not describe the same object, and until the difference is reconstructed line by line from the source data, any statement of convergence remains ambiguous. The reconciliation obeys a closure test: the candidate explanations — scope (all-causes net versus strictly monetary core), extended channels included in the consolidated net but outside the core, global overlap discount, non-monetary offsets, and the reconstruction bridge between the recomputed monetary core and the published core — must be quantified until their sum equals the measured gap (434 million TND at the low bound, 3,502 million at the central point, 8,376 million at the high bound). An unexplained residual would block the publication of any statement of convergence. This reconciliation has now been discharged (Table I3): the gap decomposes in full into ten quantified and sourced items from the case workbook, whose sum equals the gap at all three bounds — the closing residual is nil. The recomputed strictly monetary core (≈ 0.9 billion euros) brackets the published core (0.76–1.0), so that the statement of convergence of §F.16 is established under the inherited counterfactual c_V5. Two methodological reservations bound the reach of this result: the reconstruction bridge carries the bulk of the gap at the bounds — it falls to about 6% of the core at the central point — and the overlap discount remains treated as a global convention, its allocation into primitive components falling to Configuration M. The line-by-line detail and the closure test are held in the evidence file of the case.¶ 15

I.6 The Configuration M programme of work

The computation that discharges §F.16 is defined by Configuration M of Annex F (§F.15): identical stopping rules, restricted scope. Table I4 lists the deliverables and the data each of them requires. None of these steps presupposes any new theory; each is a task of data assembly and documentation.¶ 16

I.7 Bridge to Configuration F, to Annex G and to the pilot pathway

Configuration M outputs serve measurement, publication and scenario design; they can never trigger a financial consequence, and GFEA accepts no referral derived from them (§5.4). The move to regime level follows Annex G: identified Federal Reserve and ECB shocks with a joint model isolating the common global component; exposures fixed at E⁰ or instrumented; the three families of counterfactuals; interactions between channels; placebo and pre-trend tests; and the inclusion of Tunisia as one case among others in a regional panel of recipient economies with heterogeneous exchange-rate regimes, capital controls and debt structures — the second-country commitment of Annex G (regional panel), which is also the structural answer to the positional disclosure of I.1. Within the adoption pathway of Section 12.5, the Tunisian case file is simultaneously a pilot candidate for a national HuIB platform operating, at entry, under Configuration M, with the architecture’s assistance arrangements financing the move up to Configuration F.¶ 17

I.8 What this case permits and forbids one to say

The stopping rules of §F.13, applied to this case file, produce the following communication discipline. It is permitted to say that an accounting scenario classified under Annex F, under a single inherited counterfactual, places the direct external cash-flow incidence borne by the Tunisian Treasury in a range of 1.0 to 4.8 billion TND (base 2.3), rising to 1.6–6.9 billion TND (base 3.5) when partially attributed domestic transmission is included; that potential gains for the central bank and for households in receipt of transfers exist on a comparable order of magnitude and are reported separately by beneficiary; and that the Annex F framework brackets, without confirming, the monetary core reported in §9.3. It is not permitted to say that foreign central banks cost Tunisia any amount whatsoever; that a welfare loss, a causal effect, a fault or a share of responsibility has been established; that the counterparts cancel the losses, or the losses the counterparts; nor that any of these figures grounds a price, a levy or a claim. This annex is the citation reference for the case: any use of the Tunisian figures elsewhere in the document or in derived communications conforms to the present section.¶ 18

I.9 Limits specific to the case

Beyond the general limits of Section 14, five are specific to this case file. First, single-country evidence: nothing here generalises as long as the regional panel does not exist. Second, the author’s position (I.1): mitigated by structure and by external replication, not eliminated. Third, the inherited counterfactual: every figure depends on c_V5, whose polar character is declared but not yet bracketed by an alternative trajectory. Fourth, the unaudited overlaps between the extended private channels (6/7/8), which cap the evidentiary weight of the extended consolidation at the level of illustration. Fifth, the missing favourable component: as long as C3-price is not quantified, the case file is asymmetric in the unfavourable direction by construction — a conservatism that must be declared rather than silently savoured, for it too is a bias.¶ 19

@incollection{yaiche_hc_annexe_i,
  author    = {Yaïche, Nizar},
  title     = {Annex I — The Tunisian case file: evidentiary status, classification under Annex F and Configuration M trajectory},
  booktitle = {The Humanity Compact},
  year      = {2026},
  version   = {2.3},
  doi       = {10.5281/zenodo.22746533},
  publisher = {Zenodo},
  url       = {https://humanitycompact.org/ouvrage/annexe-i/}
}

Ouvrage de Nizar Yaïche, ancien ministre des Finances de la République tunisienne. Qui est l’auteur ?