Technical paper · Public Policy · Azores

From Territorial Continuity to the Economic Architecture of Public Policy

A preliminary economic analysis of the approved tender specifications and the announced public compensation. The privatisation of SATA Handling as a case study

Executive summary

The privatisation of SATA Handling is no longer an open question. The tender specifications were approved on 28 July 2026¹, the disposal of the entire share capital was decided on 22 May², and a Public Service Obligation for the activity has been announced with annual compensation estimated at around six million euros³. What remains open is the economic architecture that will, or will not, give that package verifiable content.

This document neither defends nor opposes the transaction. It proposes a method for assessing it, and applies that method to the publicly available information.

The thesis is this: privatising the company is not the same as contracting for the function. The State may transfer ownership, the assets and part of the commercial risk; it cannot transfer to the market the definition of the public interest. Territorial continuity is secured only if the obligation is defined as measurable capacity, if compensation corresponds to the efficient incremental cost of that obligation rather than to the accounting deficit, and if some body holds the mandate to calculate, verify and sanction.

Three gaps concentrate the risk in what is currently known.

The territorial obligation is framed in qualitative terms — a continuous, efficient operational response adjusted to requirements⁴ — rather than as measurable capacity station by station. An obligation whose standard is set by circumstance is not open to verification and, consequently, not open to sanction.

The announced compensation approaches the accounting loss: six million euros a year corresponds to roughly ninety-six per cent of the estimated 2025 loss of 6.25 million⁵. The proximity may be coincidental, but it requires a demonstration of which of two methods was used — the efficient incremental cost of identified obligations, or a projection of the operating deficit.

The cost benchmark is being requested from bidders rather than established by the State. The specifications require candidates to submit a forecast tariff and cost-structure model for each airport⁶ — necessary information, rightly requested — without any independent public benchmark against which those submissions can be assessed.

To these are added two structural questions that have received less attention: the buyer acquires, in all probability and subject to empirical confirmation, the position of sole supplier of an essential service at stations where there is no effective contestability; and the disposal is proceeding in parallel with the privatisation of Azores Airlines⁷, the company's dominant customer, with no publicly known intra-group terms fixed and common to both procedures.

The recommendations are set out in the final section, ordered by deadline. None requires suspending the process. All are intended to complete it.

Principal conclusions

Preliminary and methodological note

This document was closed on 30 July 2026, two days after the tender specifications were approved by the Regional Council of Government of the Azores. It is not a proposal about future specifications; it is an economic assessment of specifications already approved and of compensation already announced.

References to the content of the specifications rest on information made public on 28 and 29 July 2026, including the document to which the Lusa news agency had access, and on the questions and replies processed in the Legislative Assembly of the Autonomous Region of the Azores; they do not substitute for consulting the full document, which was not accessible at the date of closing. Where this text identifies absences, these are to be read as "absent from the disclosed information", not "absent from the approved document". No legal conclusion is formulated here: the characterisation of the compensation under the Altmark conditions, the package governing services of general economic interest and the perimeter of the 2022 restructuring decision remains pending independent opinion.

The text distinguishes four epistemic statuses. Documented: the disclosed content of the specifications, the loss series, the cost of the activity in 2025, the indicative compensation figure, the guarantee granted and the timetable of both procedures. Inference: the reading that compensation may have been anchored in the historical deficit, and the probable characterisation of sole-supplier status. Identified risk: overcompensation, rent extraction through tariffs, cross-mispricing of the two privatisations and the dilution of obligations through subcontracting. And dependent on information not yet available: the quantification of efficient cost, the causal decomposition of the deficit and the legal characterisation of the compensation.

Also noted is a divergence between sources as to the period for compliance with collective labour regulation — thirty or thirty-six months⁸ — to be confirmed against the approved document.

What an economic architecture of public policy is

The expression runs through this document and requires definition, because everything that follows depends on it.

An economic architecture of public policy is the set of verifiable elements that must exist for public money to purchase a determinate outcome, rather than to cover an outcome once recorded. It is neither a financing model nor a choice between public and private ownership: it is the structure that makes it possible to say, before paying, what is being bought, and to demonstrate, after paying, that it was delivered.

It comprises ten elements. An obligation defined as measurable capacity, not as intention. An efficient cost benchmark, established for a diligent operator under the actual conditions of the territory rather than derived from the company's own history. A causality test separating the cost of the obligation from commercial cost, from inefficiency and from the general conditions of doing business. An efficient revenue figure, so that compensation does not absorb deficient commercial policy. A reasonable return, stated ex ante. A technical ceiling limiting budgetary exposure. A settlement formula with both a floor and a cap. Cost accounting that allows the cause of every euro to be reconstructed. An audit of economic causality, distinct from an audit of accounting compliance. And an automatic rule for sharing efficiency gains, so that improvement is neither wholly confiscated nor wholly appropriated.

To these must be added an institutional condition without which the ten are inoperative: someone must hold the mandate to calculate, verify and sanction, and that body cannot coincide entirely with the party that sells the asset, defines the obligation and pays the compensation.

The order matters as much as the list. The obligation is defined before it is costed; the cost is measured before it is compensated; performance is verified before the contract is renewed. Reverse that sequence — if cost is established before the obligation exists, or if compensation is fixed before cost is audited — and the mechanism ceases to be an architecture and becomes a transfer with retrospective justification. The absence of any single element produces the same effect: the State pays for an accounting result instead of contracting for a public one.

Qualifying the figures used

Before using the disclosed figures, it is necessary to delimit what they measure. The validity of any benchmark built upon them depends on this precaution.

The figures are described as estimated losses⁹, without it being known whether they correspond to net result, operating result, result before depreciation, amortisation and provisions, or cash flow, nor how internally allocated costs, group overheads, equipment depreciation, provisions or financial charges were treated. The question is the more relevant because the activity was carved out only recently: until April 2026, handling was an internal unit of SATA Air Açores¹⁰. The historical results therefore depend on internal allocation criteria between the unit and the rest of the group — criteria that are not public and which, if altered, alter the result without the real operation changing at all.

The Regional Government itself characterised the figures as a historical accounting portrait of the activity¹¹. The observation is correct and its consequences are broader than may have been intended: a historical accounting portrait is precisely what should not serve as the basis for public compensation, and to acknowledge this is to acknowledge the need for an efficient cost benchmark. Different bases produce different benchmarks — a deficit measured before depreciation and group charges may be substantially lower than the net result — so publication of the basis of calculation is a precondition of any discussion of the amount.

The figures now public

For months the debate proceeded without a quantified basis. That changed on 29 July 2026, when the Regional Government, replying to a parliamentary request, disclosed the series of results¹².

Estimated losses were 3.45 million euros in 2021, 2.18 million in 2022, 4.25 million in 2023, 4.59 million in 2024 and 6.25 million in 2025 — a deterioration of 1.66 million on the previous year and the highest level of the period. The cost of the groundhandling service reached 35.3 million euros in 2025¹³.

A third element, earlier and decisive, must be added. On 6 July, disclosure was requested in the Regional Legislative Assembly of the studies underpinning the Regional Government's announced intention to create a Public Service Obligation for handling, with annual compensation estimated at around six million euros¹⁴. The request invoked the sector's growing public charges and pointed to a contradiction: studies commissioned by the Government itself were said to point to positive results in the operation. The Regional Government subsequently confirmed that a financial execution report for the PSO projected profit in the first three years¹⁵.

These figures do not illustrate the analysis. They determine it. And they permit three readings, stated with the degree of certainty the information allows.

First reading: the announced compensation approaches the accounting loss. Six million euros a year corresponds to roughly ninety-six per cent of the estimated 2025 loss. The coincidence may be fortuitous, or it may indicate that the figure was anchored in the activity's historical result rather than in the efficient incremental cost of a defined territorial obligation. The question should be answered with documents: does the six million figure derive from a calculation of efficient incremental cost, or from a projection of the operating deficit?

Second reading: the deterioration in the loss cannot be explained by merely invoking insularity. Between 2022 and 2025 the loss multiplied by 2.87. Although the geography of the archipelago is structurally constant, the financial cost of its consequences is not: it can vary with the volume and time distribution of movements, the composition of the fleet handled, the incidence of operational irregularities, factor prices and wages in particular, safety requirements, the cost of transporting parts and equipment, the intensity of redundancy required and the loss of density economies. Insularity may be structural while its cost varies over time. What the trajectory demonstrates is not that the territory is irrelevant, but that the growth of the deficit cannot be attributed to it automatically. Causal decomposition is required, and the burden of that demonstration falls on whoever invokes the territorial explanation.

Third reading: projected profit alongside simultaneous compensation constitutes a material risk of overcompensation. If a financial execution report projects profit in the first three years under a PSO regime, awarding six million euros a year at the same time requires documentary reconciliation between the two calculations. This is not, in itself, proof of overcompensation: the outcome depends on whether the projected profit is struck before or after compensation; whether it covers all stations or only some; whether it refers to the perimeter of the company or only to that of the public service obligation; what reasonable return was assumed; what investment, depreciation and provisions were included; and what counterfactual was used. None of these elements is public. Until they are, the coexistence of two statements by the same decision-maker — the activity will be profitable; the activity requires six million a year — constitutes a risk that the third Altmark condition obliges one to address¹⁶.

Company, function and public responsibility

A company is a legal and economic organisation. Groundhandling is an operational function. Territorial continuity is a public responsibility. The three are connected and are not the same thing — and privatisation transfers only the first.

In island territories, handling forms part of a production chain for mobility — aircraft, crew, fuel, navigation, airport infrastructure, security, ground assistance, systems, cargo and operational coordination — in which the absence of one link can render the rest useless, and an air service can exist in the timetable and disappear in reality. If the State secures an affordable fare but there is insufficient handling on an island, the policy fails; if the cost of assistance makes the route unviable, the policy fails; if there is no reserve equipment or capacity to respond to irregularities, the policy fails.

Handling is therefore an economic infrastructure, even if not a physical one: like a runway or an electricity grid, it represents installed capacity that enables other activities, with systemic value and whose absence produces costs extending well beyond the company — cancellations, loss of connectivity, effects on tourism and the widening of inequalities between islands. This does not imply that it must remain public; it implies that its privatisation requires an explicit economic definition of the territorial function that will continue to be required.

The alternative is to rely on continuity by inertia — to presume that the buyer will maintain operations because they already exist and will serve every island because the company did so before. A rational buyer will tend to concentrate resources where demand, margins and predictability are greatest, and those decisions may be entirely legitimate and even necessary to recover productivity. But what is rational for a company is not necessarily sufficient for an archipelago, because territorial continuity is not measured by average demand: it is measured by the capacity to respond to concentrated schedules, irregularities, diverted flights, emergencies and seasonal peaks. An operation may have low average utilisation and high marginal importance; it may appear redundant until the day it becomes necessary. Continuity lives on this paradox — it requires that certain resources be available before there is demand to remunerate them.

Hence minimum capacity cannot be defined by the operator after privatisation. It must be established by the public authority, in measurable terms, wherever the market offers no sufficient guarantee that it will be provided on the operator's own initiative. The right question is not who will buy SATA Handling, but what that buyer will be obliged to guarantee when commercial logic points elsewhere — and how compliance will be verified.

The context of the transaction is precise. On 22 May 2026 the Regional Council of Government decided on the disposal of one hundred per cent of the share capital¹⁷; on 28 July it approved the tender specifications, by private negotiation in two phases, and simultaneously granted a 55 million euro guarantee to the SATA group¹⁸; and in parallel the privatisation of at least seventy-five per cent of Azores Airlines is under way¹⁹.

The error of conflating loss with insularity

The disclosure of growing losses invites a politically comfortable conclusion: the company loses money because it operates in a dispersed archipelago, therefore the loss represents the cost of insularity.

The conclusion contains a measure of truth and cannot be accepted without decomposition. An accounting loss is the net result of costs, revenues, productivity, contracts, prices, organisation, asset utilisation and territorial factors; it does not, in itself, identify the economic cause of the losses. Not every additional cost of an island company is a cost of insularity; not every cost of insularity is compensable; and not every loss is a territorial cost.

If the State accepts loss as an automatic measure of territorial continuity, it will simultaneously finance what the territory makes unavoidable and what management could correct, eliminating the incentive to productivity. If it refuses any compensation, it may require structurally loss-making capacity for reasons of public interest without recognising the incremental cost of that requirement — and the service will end up reduced or abandoned.

The solution is to decompose the operation into four categories.

Normal commercial cost is what an efficient operator would bear in a comparable operation: wages, training, maintenance, insurance, systems, management and equipment.

Management inefficiency comprises costs above those of a comparable and reasonably efficient operator, without territorial justification: inadequate sizing, poor shift distribution, low equipment utilisation, deficient maintenance or badly negotiated contracts.

Structural costs of insularity are unavoidable, efficient and verifiable additional costs associated with geographical dispersion, the small size of markets, the difficulty of sharing resources between airports, the need for local crews, the transport of parts, seasonality and the reduced scope for economies of scale and density.

The incremental cost of the territorial obligation is not the whole difference between operating on the mainland and operating in the Azores, but the additional cost of securing capacity, quality or coverage that the market would not guarantee absent a public obligation.

It is this last category that can justify specific compensation. Not the overall loss, not the undifferentiated sum of expenses, not the difference between actual revenues and historical costs: only the efficient incremental cost causally attributable to a clearly defined obligation. The causality test is direct — absent the obligation, would this cost disappear or fall significantly? If the answer is no, the cost does not belong in the compensation. It is the most important defence against overcompensation, because it forces every euro to be related to a concrete obligation instead of accepting the company's overall result.

Applied to the known figures, the requirement is concrete. Of the 6.25 million euro loss in 2025, equivalent to roughly 17.7 per cent of a 35.3 million cost base, what share belongs to each category? Until that decomposition is published, qualified as to its accounting basis and audited, any compensation lacks a verifiable economic foundation.

Operationally, the decomposition translates into an allocation rule by station: the activity an efficient operator would provide absent any public obligation is financed by market revenues; additional commercial capacity chosen by the operator remains at its own risk; the efficient structural costs of insularity are recognised in the assessment but are not automatically compensable, depending on whether they follow from a public obligation or from the general conditions of operating in the archipelago; and only the extension imposed by the public interest — presence at stations, availability, reserve capacity or service levels that a purely commercial operator would not secure on the same terms — forms the core of any compensation. The mapping must start from the real operation, not from the accounting structure of a unit carved out a few months ago.

The efficient cost benchmark

The economic architecture of a public policy begins by replacing historical cost with an efficient cost benchmark. The first tells us what the company spent; the second seeks to determine what a diligent operator, adequately organised and subject to the actual conditions of the operation, ought to spend. A policy based on historical cost perpetuates the existing structure; based on efficient cost, it allows territorial specificities to be recognised without turning the company's past into a permanent financial obligation.

The calculation must be granular enough to distinguish airports, periods, services, carriers, traffic volumes and reserve capacity. An aggregate average across the nine airports would be administratively simple and economically misleading: an airport with regular, diversified traffic does not face the cost structure of a station with few movements, high hourly concentration and no possibility of sharing resources.

Comparisons with mainland operators help measure productivity and equipment utilisation, but cannot be imported mechanically: a mainland operator moves crews between nearby airports and turns to alternative suppliers within hours, whereas in an archipelago distance turns small redundancies into conditions of resilience. The objective is not to imagine a mainland operator administratively placed in the Azores, but to construct the benchmark of an efficient operator working under Azorean conditions — avoiding both the acceptance of all local expenditure as unavoidable and the treatment as inefficient of everything exceeding the cost of a denser market. Territorial efficiency does not entail eliminating all idle capacity, because in certain activities that capacity is the price of availability. The criterion is efficient necessity, not permanent utilisation: would a prudent operator, obliged to meet the defined service level, maintain that resource, in that place, in that quantity? If so, the cost may be territorially justifiable; if not, it is inefficiency, commercial overcapacity or a business decision.

Efficient cost as a transition path

The specifications impose defined restrictions: a prohibition on collective dismissals and on the elimination of posts for thirty months, compliance with collective labour regulation for a minimum period²⁰, maintenance of the registered office in the Azores for at least thirty months, and an obligation to require subcontractors to comply fully with the applicable collective labour rules.

These conditions bear directly on the calculation of compensation. For two and a half to three years, one avenue for adjusting the cost structure is unavailable. The efficient cost benchmark cannot therefore be a single level applicable from day one: it must be a path, with an initial value close to the inherited structure and convergence in defined stages towards the cost of an efficient operator under Azorean conditions, articulated with the expiry of the restrictions.

It is important to be explicit about the scope of this observation. The labour restrictions limit one avenue of adjustment; they do not eliminate the other sources of productivity, which must be mobilised from the outset and which constitute the more substantial part of the convergence potential: reorganisation of shifts and rosters, flexible allocation of resources between services, multi-skilling, digitalisation of planning, preventive rather than corrective maintenance, renegotiation of procurement, improved hourly utilisation of equipment and revision of the tariff. The efficiency of a groundhandling operation lies above all in planning and asset utilisation, not in the nominal size of the workforce.

From this follows an explanation for the announced compensation that is economically more defensible than covering the loss. If the buyer cannot, in the short term, fully close a deficit of this order, the sale price will tend towards very low values or compensation will be required from the outset. This justifies transitional support, not permanent support calibrated on the historical loss. The distinction should appear in the contract: transitional compensation, declining and tied to the expiry of the restrictions, differs in nature, duration and legal characterisation from structural public service compensation associated with permanent territorial obligations.

The subcontractor clause confirms that subcontracting is expected. Without clauses passing the territorial obligations through in full — capacity, availability, readiness, quality and information — the obligation imposed on the operator can be progressively hollowed out by contract while remaining formally satisfied at the level of the holding entity.

Efficient revenue and the compensation formula

Debate on public services tends to concentrate on costs, but rigorous compensation also requires the calculation of efficient revenue: a company may show insufficient revenue not only because the market is small, but because prices are inadequate, contracts were badly negotiated or certain services are not invoiced. Efficient revenue represents what a diligent operator ought to obtain under the actual conditions of the market — a prudent benchmark based on volumes, comparable prices, quality, contractual conditions, seasonality and the carriers' ability to pay. Without it, the State would pay the difference between all costs and revenues actually earned, including the part resulting from inefficient commercial policy.

The admissible compensation is formulated as follows:

In this expression:

• C is the efficient incremental cost of the public service obligation, established station by station under the causality test and transition path described above;
• R is the reasonable return on the capital actually mobilised and the risk actually assumed;
• Rev is the efficient revenue attributable to the obligation;
• O is other compensation, subsidies or economic advantages received, direct or indirect;
• F is the operator's own deductible, that is, the share of the deviation it bears without compensation;
• L is the technical ceiling, set ex ante against the audited efficient cost.

In this matter the form of the expression is as important as its content. The min(·, L) operator introduces a technical ceiling: no circumstance can generate compensation above the previously audited limit, transferring to the operator the risk of overrun beyond the efficient benchmark. The max(0, ·) operator prevents negative compensation, avoiding the conversion of the mechanism into a transfer from the operator to the State through cyclical fluctuation. A formula without a ceiling and without a floor is, in practice, an open-ended commitment of the public budget.

As to the deductible F, a methodological clarification is required. The architecture of the Hybrid Territorial Continuity Mechanism includes a beneficiary's own contribution, which in that context has a precise meaning: it is the share paid by the resident at the point of sale. In handling there is no individual beneficiary bearing a share; the counterparties are the carriers, which pay a tariff, and the State, which pays compensation. The variable cannot therefore be carried across from one mechanism to the other by formal analogy. If adopted, F must have an autonomous economic foundation — typically the retention of a share of risk by the operator, serving to prevent moral hazard and preserve the incentive to efficiency — and be calibrated in that capacity. If it lacks that foundation, F equals zero and disappears from the expression. Every variable in a compensation formula must have its own operational meaning in the obligation to which it applies.

The reasonable return must be stated explicitly, not left to discretionary appraisal: denying any return would render the contract unviable or push remuneration into less transparent components. The appropriate reference is a weighted average cost of capital calculated for the activity and for the risk profile actually transferred, with an alternative safe harbour — of the relevant reference rate plus a fixed margin — where no reliable comparator can be established. What is not admissible is a return guaranteed irrespective of performance: normal commercial risk must remain, in relevant part, with the operator. The territorial obligation can be compensated; poor management cannot.

The allocation methodology must likewise be stated. A calculation based on net avoided cost — the difference between the operator's position with and without the obligation — produces results different from a cost-allocation calculation, and is as a rule preferable in order to avoid common costs being absorbed by the public obligation²¹. The choice should be declared, reasoned and kept stable over the life of the contract.

Sharing efficiency gains

A mechanism confined to covering the difference between costs and revenues eliminates the incentive to efficiency: any saving reverts to the State and any deterioration is absorbed by the budget. The corrective is an automatic rule for sharing deviations from the efficient benchmark, with three destinations — retention by the operator for a defined period, return to the State through reduced compensation, and application to improved capacity or prices.

The incentive architecture used in the hybrid SSM/Azores Tariff model and in the MCTH takes the form of a three-way split, with a calibration of sixty, thirty and ten per cent. That reference is presented here as an illustrative example of architecture, not as a transferable parameter. The percentages applicable to handling will have to result from their own calibration, for three substantive reasons: the activity has a capital intensity and fixed-cost structure different from passenger transport; the relevant demand elasticity is that of the carriers, not of residents; and the recipient of the improvement share is the air operator rather than the citizen, which changes the distributive effect of the third component.

Two adaptations merit consideration: a retention period aligned with the seven-year horizon of the projections required from bidders²², so that the operator has an incentive to investments whose return exceeds the financial year; and channelling the improvement share towards capacity and availability at lower-traffic stations, given that a general tariff reduction benefits mainly the dense routes. The principle holds whatever the percentages: an operator that reduces costs without degrading service keeps an identifiable part of the benefit, and the State recovers a part, automatically and without negotiation. The same applies to revenues.

The approved specifications: an audit in ten elements

Tender specifications should not confine themselves to defining who may buy. They must establish what the buyer will be obliged to do, in terms open to measurement, verification and sanction.

The disclosed information reveals a substantial set of requirements: suitability and a minimum of five years' relevant experience in groundhandling or in the transport, logistics or infrastructure sectors; guarantees of financial capacity; a binding strategic plan; seven-year projections with traffic scenarios; a forecast tariff and cost model by airport²³; labour restrictions and maintenance of the registered office²⁴; a prohibition on disposing of the stake for five years, with a subsequent right of first refusal²⁵; a minimum obligation of continuous, efficient operational response adjusted to requirements²⁶; and continuity of provision, in general terms, at the nine airports and aerodromes²⁷. It is a serious set of requirements in corporate, financial and labour terms. The assessment that follows addresses the remaining dimension: the economic and territorial specification of the service.

The balance is asymmetric. In corporate, financial and labour matters the specifications are detailed and binding. In the territorial and economic dimension — precisely the one that justifies public intervention and any compensation — the known obligations are qualitative and their verification is unstructured. Generic promises prove difficult to enforce when they collide with concrete business decisions.

Information asymmetry

One aspect of the specifications deserves separate treatment, because it is not a gap but an inversion.

The document requires bidders to submit a forecast tariff and cost-structure model for each airport and aerodrome, with traffic scenarios and a seven-year horizon. The requirement is sound and the information gathered is exactly what a territorial matrix needs.

The problem is the direction of flow. The State is asking candidates to tell it what the operation costs, without any known independent benchmark against which submissions can be assessed. In those conditions the bidder's model is simultaneously information about reality and an instrument of negotiation over future compensation, and a rational candidate will construct projections supporting the highest sustainable compensation, not the most accurate estimate of efficient cost. The procedure thus risks institutionalising information asymmetry at the very point where the economic architecture exists to close it.

The corrective does not require altering the timetable: establish, before binding proposals are assessed, a public efficient cost benchmark by station, with independent technical support and comparators adjusted to Azorean conditions; use the bidders' models as a cross-check on that benchmark rather than as its substitute; and set the technical ceiling L against the audited value, not the proposed one.

Without this step, compensation will be determined by the information of the party receiving it. With it, it becomes determined by the information of the party paying.

Market access, monopoly and competition

Analysis has concentrated on the risk that the operator reduces service at the least profitable stations. There is a symmetrical and equally material risk: that the buyer acquires a dominant position over a factor of production indispensable to any carrier wishing to operate in the Azores.

The specific European framework is Directive 96/67/EC on access to the groundhandling market²⁸. Its regime is graduated: the opening of third-party handling and self-handling is tied to traffic thresholds, and at airports below those thresholds the number of suppliers may be limited, with sole-supplier situations able to persist.

A methodological reservation is required: the assertion that the asset confers sole-supplier status cannot rest on legal plausibility alone, and requires empirical verification, station by station, of the elements in the table below — none of which appears in the available information.

Subject to that verification, the probable configuration raises three requirements that the specifications, as disclosed, do not appear to address.

The first is an obligation of non-discriminatory access: provision to all carriers on transparent, objective and non-discriminatory terms, with publication of reference tariffs and a prohibition on conditions favouring companies within the same group. The question is the more relevant because the buyer may come from the transport sector — a possibility expressly admitted in the experience requirements²⁹ — introducing a risk of vertical integration and self-preferencing.

The second is price supervision. A sole supplier without contestability can extract rent through the tariff without reducing capacity or breaching any continuity obligation: the service continues, the State pays the territorial compensation and the surplus is captured through the price charged to the carriers.

The third follows from the interaction with territorial continuity policy, and is the most important. If the State finances air accessibility — through the inter-island concession, through public service obligations or through a resident-fare mechanism — while allowing the groundhandling supplier to set the price of that service freely, part of the public envelope intended for continuity may be transferred to the handling operator as a station cost. The support reaches neither the resident nor the carrier: it dissipates into an unregulated intermediate link. Any assessment of a continuity mechanism that ignores the price of handling overstates the effect of the support.

It is in this context that one of the disclosed selection criteria gives rise to reservation: among the assessment factors is the maintenance and reinforcement of the disposed company's competitive position³⁰. In a market without effective contestability, reinforcing that competitive position is not a neutral objective — it may amount to consolidating a dominant position over an essential service, to the detriment of the carriers that depend on it and, indirectly, of passengers and of the public purse that finances accessibility. The criterion is understandable from the standpoint of asset value; it is problematic from the standpoint of competition and of the effectiveness of public expenditure.

Sequencing with the Azores Airlines privatisation

SATA Handling is being disposed of in parallel with the privatisation of at least seventy-five per cent of Azores Airlines³¹. Both transactions form part of the same restructuring plan and of the European Commission's June 2022 decision, which approved state aid of 453.25 million euros in loans and guarantees, providing for the divestment of a controlling stake³².

SATA Handling's dominant customer is the SATA group, which creates a sequencing problem that the known documentation does not appear to resolve. Whoever acquires the handling company will be valuing a revenue stream whose principal counterparty changes ownership within months, on terms it did not negotiate; whoever acquires the airline will inherit a handling contract concluded before its entry, with a supplier privatised in the meantime. Each buyer faces counterparty risk in relation to the other, and neither controls the instrument that mitigates it.

The consequence is economic, not merely legal. Absent intra-group terms fixed, published and identical across both processes, the State risks incorporating the same counterparty risk into the valuation of both assets, reducing the value obtained and increasing the likelihood of subsequent conflict — a risk aggravated if the same economic group positions itself in both procedures.

The corrective is practicable: fix and publish, before binding proposals in either process, the groundhandling contract between SATA Handling and the group companies — with price, scope, service levels, duration and revision rules — and include it as a binding element in both sets of specifications.

The net public value of the transaction

Debate on privatisations tends to concentrate on sale proceeds. The metric is insufficient and can mislead: a high price may come with future commitments exceeding the amount received, and a low price may be justified if the buyer assumes substantial investment, risk and obligations. The correct assessment is of net public value:

where P is the disposal price; PV(Comp) the present value of the public compensation contracted; Liab the liabilities retained by the State or the public holding company; E(Cont) the expected value of contingent risks, including guarantees; Sav the capital and operating savings arising from the transfer; and Perf the performance benefits attributable to private management, to the extent contractually enforceable and measurable.

Applied to this case, the equation has two immediate implications. The 55 million euro guarantee granted on the same day as the specifications were approved³³ belongs in the contingent risk term and must be valued, not treated as a free-standing administrative act. And annual compensation of six million represents, even without growth and at moderate discount rates, a present value of several tens of millions over a typical contract period — an order of magnitude that dominates any plausible sale price for a company with the disclosed record.

The conclusion of method is that the decision should not be assessed by the price obtained, but by the balance between what the State receives and what it undertakes to pay and to guarantee. Without this equation stated and quantified, debate will revert to the nominal value of the sale, the least informative variable in the set.

The European framework

Any public compensation associated with this activity must be analysed under the European rules on competition, public procurement, services of general economic interest and state aid.

The Altmark case law. The judgment in Case C-280/00³⁴ establishes four cumulative conditions for compensation for public service obligations not to constitute state aid: clearly defined obligations actually entrusted to the undertaking; parameters of calculation established in advance in an objective and transparent manner; compensation not exceeding what is necessary to cover the costs occasioned by discharging the obligations, account being taken of revenues and a reasonable profit; and, where the undertaking is not chosen through a procedure allowing selection of the provider able to supply the service at least cost to the community, determination of compensation on the basis of the costs of a typical undertaking, well run and adequately equipped.

The fourth condition merits particular attention. Given that a private negotiation procedure has been adopted, it is not evident, on the information known, that the first limb of that condition is satisfied. The question is not resolved by the procedure's formal designation in national law: it depends on its actual characteristics of competition, publicity and transparency, and on its capacity to select the provider able to supply the service at least cost to the community. If the procedure does not permit that demonstration, the compensation will have to be justified by reference to the costs of a typical, well-run and adequately equipped undertaking — that is, precisely through the construction of the efficient cost benchmark described in this document. To that extent, the methodology proposed here is not an analytical preference: it is, on that hypothesis, a compliance requirement. Which of the two routes applies depends on legal analysis that this document does not substitute for.

The package on services of general economic interest. Decision 2012/21/EU³⁵ exempts from notification, under certain conditions, compensation for services of general economic interest, requiring an explicit entrustment, definition of the obligations, compensation parameters, mechanisms for recovering overcompensation and periodic control. The Union Framework³⁶ sets additional requirements for situations outside the Decision, including a preference for net avoided cost and a requirement of efficiency incentives. There is also a specific de minimis regime³⁷, whose thresholds are insufficient given the orders of magnitude involved.

Directive 96/67/EC. The market access regime conditions not only competition but the characterisation of the obligation itself: the nature of any access limitation and the existence or otherwise of a sole supplier influence the definition of the service, the selection of the operator and the assessment of the economic advantage conferred.

Article 349 of the Treaty on the Functioning of the European Union³⁸ recognises the specific conditions of the outermost regions and strengthens the legitimacy of adapted measures, without removing the need for rigour: the outermost-region argument is not a dispensation from calculation, but grounds for a more appropriate calculation.

A clarification distinguishing this case from passenger transport is called for. Article 51 of the General Block Exemption Regulation³⁹ concerns social aid for transport for residents of remote regions, and that is the basis customarily invoked in schemes supporting residents' mobility. It does not cover compensation granted to a groundhandling provider. Extending the economic architecture of a territorial continuity mechanism to handling is conceptually coherent, but it does not carry the same legal basis with it: the route applicable to handling is that of services of general economic interest. This discontinuity should be flagged, not concealed.

A further element heightens the evidentiary requirement: the activity falls within the perimeter of a restructuring plan already the subject of a Commission decision in 2022⁴⁰, and a 55 million euro guarantee was granted to the group on the day the specifications were approved. Any new compensation will not be appraised in isolation, but within the framework of that decision and its conditions, including the totality of measures in favour of the group perimeter. Appraising the handling PSO in isolation, without regard to the guarantee, the inter-island concession and the earlier measures, would be methodologically incomplete.

It should not be asserted in advance that a solution falls outside the state aid regime, nor concluded that any compensation is incompatible — the answer requires independent legal qualification. The best defence will in any event be sound economic design: compensation based on efficient costs, efficient revenues, a stated return, measurable obligations, a technical ceiling and periodic audit is more robust than a transfer intended to cover overall losses.

The institutional gap

The foregoing analysis assigns to a public body a demanding set of functions: establishing the efficient cost benchmark by station, setting the technical ceiling, verifying the economic causality of declared costs, establishing efficient revenue, calculating and settling compensation, applying the gain-sharing rule, measuring service levels, imposing sanctions and conducting parametric revisions.

It is not known who holds them. And the problem is not solved by identifying a single body, because the functions are of different kinds and may legitimately be distributed. What the contract must do is assign them expressly.

The difficulty lies in the last two rows. Absent an express solution, the Regional Government would combine the positions of selling shareholder, authority defining the obligation, counterparty paying it and body verifying compliance — a concentration that compromises the credibility of the verification, whatever the technical quality of those performing it.

An economic architecture without an institutional holder is a document, not a mechanism. Three configurations are possible: reinforced competences over access and tariffs assigned to the civil aviation authority; a regional body supervising territorial continuity contracts, with a mandate covering the inter-island concession, public service obligations and handling; or outsourcing calculation and audit to an independent technical body, with mandatory publication of reports and a prior opinion before settlement. The choice is political; the absence of a choice is an implicit decision in favour of the status quo.

Procedure design and untested alternatives

Specifications that are too restrictive produce a procedure without bids; specifications that leave essential obligations vague produce formal success and economic failure after the sale. The solution is not to choose between rigour and attractiveness, but to turn uncertain obligations into calculable risks: an investor will accept maintaining capacity on lower-demand islands if it knows what the obligation is, how efficient cost will be measured, what revenues will be counted, what return is permitted and on what terms the contract may be revised. What deters serious investors is not the existence of obligations, but unpredictability.

Three design questions have remained outside the public debate.

The structure of the asset disposed of. The procedure is the outright disposal of a single company covering nine stations of very unequal profitability; the alternative would be to group stations into lots combining dense and loss-making operations. Three things that do not coincide must be separated: the aggregation of stations, the competitive design of the procedure and compliance with the fourth Altmark condition. Aggregation may improve the economic sustainability of each lot, but does not in itself satisfy that condition — only a genuinely competitive procedure could be relevant to its first limb. And dividing an already constituted company is not necessarily superior to integrated disposal: it may destroy synergies, duplicate administrative costs and reduce network economies. It should have been the subject of comparative simulation before the decision, and it is in that capacity that it is noted here.

Ownership of critical assets. Groundhandling equipment is mobile and redeployable, and nothing prevents its progressive concentration at the more profitable stations. A structural response is to retain public ownership of critical equipment at lower-traffic stations, made available to the operator under tied use, with automatic reversion on breach or termination. The buyer acquires the operation; the State retains the physical instrument of continuity. This is more effective than access rights invoked after a rupture.

The comparison of instruments. The analysis in this document is instrument-neutral: the architecture proposed does not presuppose privatisation, it is the condition for comparing options. Four alternatives merited documented assessment — partial disposal with a minority public stake and special rights; splitting the perimeter, disposing of dense stations and contracting the thin ones through a public service obligation; self-handling by carriers where technically feasible, introducing contestability without disposal; and public ownership subject to a management contract with efficiency targets, a cost ceiling and consequences for non-performance. The absence of that comparison does not invalidate the decision, but deprives it of comparative justification — and it is that justification which the parliamentary requests now demand⁴¹.

The data required

This document identifies a method; it does not calculate the correct level of compensation, and cannot do so with the information publicly available. The table below sets out the minimum data needed to convert the method into calculation — and their unavailability is itself an observation about the process, because none is accessible to anyone wishing to scrutinise the announced compensation.

Risk matrix

From the case to the architecture

Public intervention was for too long conceived as ownership or financing: the company was kept when the activity was strategic, capital was injected when losses appeared, and distance was sought when it was privatised. Economic architecture demands a different role — defining outcomes, designing incentives, allocating risks, establishing metrics, simulating scenarios, contracting capacity, auditing costs and revising decisions. Intervening less in day-to-day management and more in the quality of the model. On this view, public ownership ceases to be the only way of protecting the collective interest, and privatisation ceases to be synonymous with the State's withdrawal.

The State does not need to operate pushback tractors, steps or baggage belts. It needs to guarantee that those resources exist where they are needed, that they are used efficiently and that there is a response when the market fails. Under budgetary constraint the problem is not to spend less, but to know what is being bought with each euro: compensation without indicators buys temporary financial peace; tied to capacity, quality and outcomes, it buys territorial continuity. The same problem exists in maritime transport, energy, health and supply, where the traditional answer oscillates between historical financing and abandonment to the market. The third way is to define the obligation, calculate the efficient incremental cost, estimate efficient revenues, remunerate the appropriate risk, measure outcomes and adjust the policy — which does not guarantee politically easy decisions, but does make the choices visible.

An operational architecture developed for this purpose already exists for the continuity of air transport between the Autonomous Regions and the mainland — the Hybrid Territorial Continuity Mechanism, MCTH — whose four central elements are transferable: the efficient cost benchmark, the technical ceiling, the sharing of risk and gains, and preventive simulation with continuous evaluation. Application to handling extends the same logic to another link in the chain rather than constituting a first application of the mechanism; and the validity of the analysis presented here does not depend on adherence to that mechanism or to its parameters — it derives from the economics of regulation and from the facts of the case. It is the capacity for simulation that was missing: the decision-maker ought to have known, before approving the specifications, what it costs to require permanent presence on every island, what a partial-availability solution costs, what interruption risk follows from reducing a given item of equipment, and what effect different traffic scenarios have on the compensation due.

One final distinction, with direct practical consequence. Territorial continuity secures the link between parts of the national territory; strategic international connectivity serves tourism, investment, exports or integration into global networks, and does not follow from the same internal obligation — it is the domain of the Hybrid International and Strategic Territorial Connectivity Mechanism, MCTHIE. Here the distinction matters because the installed capacity simultaneously serves inter-island, mainland and international services: cost accounting must prevent international commercial development costs from being charged to internal territorial continuity, and the revenues from international connectivity, which improve the utilisation of equipment and crews, must be taken into account in the calculation, on pain of the State financing capacity already remunerated by the market.

Recommendations, by deadline

Conclusion

The State may privatise SATA Handling. It may not privatise territorial continuity. It may cease to be a shareholder without ceasing to define the public interest, and transfer management without relinquishing regulation, measurement and the capacity to intervene. What it may not do is presume that the market will spontaneously produce the territorial coverage that citizenship requires, nor that a transfer calibrated on the historical deficit constitutes a continuity policy.

The central question is not whether the company is public or private, but whether the model produces the intended public outcome at the lowest efficient cost, with an appropriate allocation of risk and full transparency: from ownership as a substitute for policy to the intelligent contracting of outcomes; from loss as an argument to the analysis of its causes; from undifferentiated compensation to verifiable incremental cost. The thesis does not depend on resolving the uncertainties noted here — it depends only on accepting that a public obligation must be defined before it is paid for, measured before it is compensated and verified before it is renewed. That is why this process deserves attention well beyond the airport sector: to privatise without architecture is to change the owner; to privatise with architecture may change the quality of the State.