From Aid Dependency to Generative Reciprocity - Rethinking Development Assistance in a Relational Framework 【（Preliminary）Draft】

Abstract

The debate over dependency in development assistance is stalled for want of a criterion. Practice doctrine holds an aspiration, expressed as capacity building, ownership, sustainability, and exit, and supplies no operation by which a given arrangement is assigned to one side of the distinction that aspiration invokes. This paper supplies such an operation. It first establishes that the general dependency claim stands without support in the randomised and review evidence, so that dependency is treated as a property of the design of an arrangement, and it concedes that the distinction between maintaining and capacitating support is already held in five independent literatures, none of which supplies a test. The test proposed is the fading test, drawn from the fading component of instructional scaffolding: a support is generative where its withdrawal leaves the supported relation intact, and extractive where its withdrawal collapses the relation it sustained. Three features of an arrangement are shown to settle the result, namely the routing of the coupling a transfer creates, the visibility of the transfer to the supported party, and the position the arrangement assigns to that party, while the volume of the transfer, its medium, and the supplier’s intention stand inert. The prevalence of extractive arrangements is then explained through the asymmetric enforcement of accountability in the assistance sector, which moves each of the three determinants in the extractive direction. One axis of return, arriving at a level of organisation above that of the transfer, is entered as an observation. The paper closes by establishing the constraint under which the criterion is employed: it classifies arrangements, confers worth on no party, and supplies grounds for redesign in place of grounds for allocation.

Keywords: development assistance; aid dependency; capacity building; accountability; relational ethics

1 Introduction

The role of this section is introductory. It sets out the practical and doctrinal background against which the paper is written, states the impasse the dependency debate has reached and the motivation arising from it, formulates the questions the paper discusses, states in preliminary form the insights the body of the paper establishes, and describes the procedure by which the argument proceeds. The method of the section is expository, and the warrant for each preliminary insight is supplied at the section indicated alongside it.

Development cooperation operates today under a doctrine whose central terms were fixed across two decades of high-level agreement. The Paris Declaration on Aid Effectiveness placed ownership first among its principles, committing partner countries to exercise leadership over their development policies and donors to align behind that leadership, and the Accra Agenda for Action and the Busan Partnership carried the commitment forward while adding capacity development and the use of country systems [doctrine] (Organisation for Economic Co-operation and Development, 2005; Fourth High Level Forum on Aid Effectiveness, 2011). The same period established sustainability and exit as standing dimensions of programme evaluation, so that an intervention is assessed by the persistence of its effects after cooperation ends. Japanese cooperation had fixed the corresponding principle earlier still, directing assistance toward the self-help effort of partner countries and evaluating projects by their capacity to develop on their own footing thereafter [doctrine] (Government of Japan, 1992; Ministry of Foreign Affairs of Japan, 2015).

The doctrine holds an aspiration common to several literatures standing at some distance from one another. Welfare policy in Europe organised a comparable aspiration under the heading of social investment, treating expenditure on capacitation as distinguishable from expenditure maintaining income (Morel et al., 2012; Hemerijck, 2013).

Humanitarian relief carries the aspiration in negative form, through a standing anxiety that relief supplied at adequate levels erodes the recipients’ own initiative (Harvey and Lind, 2005). Social psychology carries it in experimental form, through the separation of help supplying complete solutions from help supplying tools (Nadler, 2002). Enterprise design carries it through the separation of extractive from generative ownership (Kelly, 2012). Human security doctrine carries it through the pairing of protection with empowerment (Commission on Human Security, 2003). The recurrence across settings this varied indicates that the underlying distinction answers to something practitioners in each setting encounter.

The debate arising from this aspiration is conducted under the heading of dependency, and it has proved durable. One side holds that assistance sustained over long periods produces the incapacity it addresses, and draws the practical conclusion that assistance should be time-bound and conditioned. The other side holds that the dependency anxiety operates as a rationalisation for supplying assistance below assessed need, and draws the practical conclusion that the anxiety should be set aside. Both sides have held their positions for decades, and the evidence assembled over the same period has settled the dispute at the level of transfers as a class while leaving it entirely open at the level of arrangements taken one at a time.

The impasse arising from this configuration has a specific form, and the motivation for the present paper lies in that form. The defect is the absence of an operation.

Practice doctrine holds an aspiration and supplies criteria of assessment applicable at completion, and it leaves open the question of how an arrangement in operation is assigned to one side of the distinction the aspiration invokes. A programme officer holding the doctrine and facing a live portfolio has available the volume of each transfer, the instrument through which it flows, the stated purpose of each intervention, and the outcome indicators reported against it. The doctrine supplies the aspiration that the portfolio should build capacity, and the officer’s assignment of any given arrangement to the capacitating or the maintaining side proceeds by judgement that the doctrine leaves unguided.

Three consequences follow, and each is visible in the literatures surveyed in this paper. Assessment migrates to the point of completion, since the evaluation frameworks that do address persistence operate ex post, which places the finding beyond the reach of the design it would inform. Design guidance remains exhortative, since a doctrine identifying an objective without identifying the features that produce it can recommend the objective and leave the features to be discovered case by case. And the vocabulary of dependency, lacking an operation of its own, remains available for allocative use, so that the description of a recipient as dependent performs work in decisions about the level of assistance supplied. These consequences persist because the missing element is a definition in place of a finding. Additional evaluations of completed programmes accumulate without supplying the operation, since each such evaluation presupposes a classification of the arrangement it assesses. The situation calls for a criterion, and the paper’s undertaking is to supply one at a scope narrow enough to be defensible.

Five questions organise that undertaking. The first concerns the operation itself:

given a support arrangement in operation, by what test is that arrangement assigned to the extractive or to the generative kind, in a manner applicable before completion and taking as its object the relation the arrangement maintains? Section 4 answers it. The second concerns the features settling the result: which properties of a support arrangement determine the assignment the test returns, and which properties, presently used to discriminate among arrangements in practice, stand inert with respect to it? Section 5 answers it. The third concerns prevalence: given a sector holding the capacitating aspiration in its own doctrine, what accounts for the systematic production of arrangements of the maintaining kind? Section 6 answers it. The fourth concerns return: in what forms does the value generated within a supported relation return, and which of those forms are already accounted for by existing theories of reciprocity? Section 7 answers it. The fifth concerns use: given that a classification of support arrangements stands adjacent to decisions about the allocation of support, under what constraint may the criterion be employed so that the classification performs diagnostic work alone? Section 8 answers it.

The insights the paper establishes are stated here in preliminary form, each with the section supplying its warrant. Their statement in advance serves the paper’s structure, since a reader holding them can follow the concessions of Section 3 as concessions in place of retreats.

Dependency is a property of design. The proposition that transfers as a class produce durable reductions in recipients’ own effort stands without support in the randomised and review evidence, and the variation the same evidence reports across designs locates the phenomenon in the structure of arrangements (Section 2). The distinction the phenomenon invokes is already held, and the operation is what remains open: five independent literatures name the separation between support maintaining a recipient’s position and support building a recipient’s capacity, one of them supports it experimentally, and each supplies an operation of assignment serving its own purposes, requiring either a completed intervention or an inspectable design document (Section 3).

The fading test supplies the operation. A support is generative where its withdrawal leaves the supported relation intact, and extractive where its withdrawal collapses the relation it sustained. The test originates in the fading component of instructional scaffolding, and the transfer into assistance settings carries the operation of staged withdrawal while leaving behind the instructional presupposition of a trajectory toward self-regulation (Section 4). Kind is settled by three features of an arrangement: the identity of the relation a transfer strengthens, the salience of the transfer to the supported party, and the position the arrangement assigns to that party. The volume of the transfer, the medium through which it flows, and the purposes of the supplying party stand inert with respect to it (Section 5).

Extractive drift has a structural explanation. Accountability in the assistance sector runs upward toward the parties supplying funds and remains weakly enforced toward the parties whose capacity the arrangements address, so the feedback loop that would reward capacity creation is broken at its return path, and benevolent purpose coexists with extractive design for that reason (Section 6). Return travels along three axes, of which two are already held: delayed and upstream return and return arriving from third parties are covered by the theory of indirect reciprocity, while return arriving at a level of organisation different from the level at which the transfer was made stands as the axis on which this paper enters an addition, and the addition is entered as an observation without formal treatment (Section 7).

The criterion classifies and confers no worth. Worth is grounded constitutively and prior to any criterion, on the ground that a thing may hold final value whose source lies outside itself. Protection of dignity holds lexical priority over every generative accounting, so that the classification of an arrangement supplies grounds for redesign and supplies no grounds for the withdrawal of support from the parties an arrangement covers (Section 8).

The paper proceeds by concession, definition, determination, explanation, and constraint, in that order, and the ordering is deliberate: the criterion is stated after the territory already occupied has been mapped, and its use is constrained before its design implications are drawn. Section 2 fixes the evidentiary standing of the dependency claim and states the reformulation the paper adopts. Section 3 exhibits the five prior holders of the distinction and identifies the operation they leave open. Section 4 states the criterion, traces its provenance, bounds the transfer from that provenance, and separates the criterion from outcome, duration, and expressed satisfaction. Section 5 establishes the three determinants of kind and the inertness of three further features.

Section 6 supplies the structural explanation of extractive drift. Section 7 treats the axes of return and states the cross-scale claim together with its limit. Section 8 states the ethical constraint governing the criterion’s use. Section 9 draws the implications for the design of assistance. Section 10 states the limits of the account and the observations that would tell against it. Section 11 summarises the results of the paper.

Two conventions hold throughout and are stated here so that they require no restatement at each occurrence. Every claim carried from another literature is labelled by its type of evidence, as formally modelled, as empirically tested with a statement of strength, or as doctrine, since the assistance literatures cite the three in a single register and the labelling is the paper’s principal defence against inheriting a strength its sources lack. And every result obtained at one level of organisation is held at that level until an explicit argument transfers it, since the evidence assembled here runs from dyadic interpersonal support through programme evaluation to inter-state cooperation, and the transfer among those levels is an argument to be made in place of an assumption to be relied upon.

2 The Empirical Standing of the Dependency Claim

The role of this section within the paper is to fix the evidentiary status of the proposition that assistance produces dependency, since the criterion developed in Section 4 would otherwise inherit whatever standing that proposition holds. Its objectives are two. The first is to report what the assembled evidence supports and what it leaves open, separating the general claim about transfers from claims about particular designs.

The second is to state the reformulation the paper adopts, under which dependency is treated as a property of the design and incentive structure of an arrangement. The section proceeds in five parts. Section 2.1 states the claim in the forms in which it circulates. Section 2.2 reports the humanitarian relief evidence. Section 2.3 reports the cash transfer evidence. Section 2.4 treats the political function the dependency label performs and the distance this paper holds from it. Section 2.5 states the reformulation, and Section 2.6 summarises the results. The method is review of the assembled quantitative and qualitative evidence, with each body of evidence labelled by its type and the strength of its support stated.

2.1 The Claim in the Forms in Which It Circulates

The dependency claim circulates in three forms of markedly different strength, and the assistance literatures move among them within single documents. The strong form holds that transfers to a household, community, or state produce a durable reduction in that recipient’s own productive effort, so that transfers generate the condition they address. The moderate form holds that transfers displace local provision, so that markets, kin networks, or public administrations withdraw from functions the transfer covers. The weak form holds that particular designs, distinguished by their targeting rules, duration signals, and conditionality, produce reductions in effort or displacements of provision, while transfers as such carry no such tendency.

The three forms carry distinct evidentiary burdens, and the sections following establish that the assembled evidence bears on the strong form directly and supports the weak form. The distinction matters for the present paper, since a criterion of support built atop the strong form would apply to transfers as a class, and a criterion built atop the weak form applies to arrangements individually. The paper builds atop the weak form.

2.2 The Evidence from Humanitarian Relief

Harvey and Lind conducted the principal review of dependency in humanitarian relief for the Humanitarian Policy Group, examining the concept’s use in relief discourse alongside the field evidence supporting it (Harvey and Lind, 2005). Their findings run in three directions. Recipients of relief were found to draw on it as one element among several within diversified survival strategies, so that reliance on relief stood well below the level relief discourse assumes. The evidence that relief erodes recipients’ own initiative was found to be thin, and the fear of such erosion was found to rest on assertion in place of demonstration. And the practical consequence of the fear was found to run toward the supply of assistance at levels below assessed need, on grounds of avoiding a dependency the evidence leaves unestablished [empirically tested: review of field evidence with case material; qualitative, with the assembled base weighted toward East and Central Africa].

The finding carries a caution about its own reach, which the review states. Absence of demonstrated erosion across the assembled cases leaves open the performance of particular designs, and the review’s own recommendations concern design features governing predictability, targeting, and exit. The evidence therefore supports the weak form of the claim and holds the strong form without support.

2.3 The Evidence from Cash Transfers

Cash transfers supply the strongest evidentiary base bearing on the dependency claim, since they have been evaluated by randomised design across a wide range of settings and their effects on labour supply and expenditure composition have been the object of dedicated synthesis.

Banerjee, Hanna, Kreindler and Olken reanalysed seven randomised controlled trials of government-run cash transfer programmes across six countries, harmonising the data definitions and the empirical strategy so that the estimates would be comparable, and examining effects on adult labour supply. Across the seven programmes the reanalysis found no observable effect on the propensity to work or on hours worked, for men or for women, and pooling across the five comparably designed studies returned the same result at higher power [empirically tested: reanalysis of seven randomised trials under harmonised definitions; the null result is consistent across the programmes examined and its reach is bounded by the six countries covered] (Banerjee et al., 2017). Evans and Popova synthesised the evidence on expenditure composition, addressing the claim that transfers flow toward alcohol and tobacco.

The synthesis found effects on such expenditure clustered at or below zero across the assembled studies [empirically tested: systematic review with meta-analysis of quantitative studies; the direction of the result is consistent, with individual studies varying in precision] (Evans and Popova, 2017). Bastagli and colleagues assembled the broader review, covering fifteen years of cash transfer evaluations across outcome domains and reporting the design features associated with variation in effect [empirically tested: systematic review of a large evaluation base; findings vary by outcome domain and by design] (Bastagli et al., 2016).

Two properties of this evidence bear on the present paper. Its strength stands high by the standards of the assistance literatures, since randomised assignment and common-specification reanalysis address the selection problems that beset observational work on transfers. And its object is the strong form of the dependency claim, which the evidence leaves without support. Variation across designs, reported in the Bastagli review, supplies the positive finding this paper builds upon: the effects of a transfer track the features of the arrangement through which it flows.

2.4 The Political Function of the Dependency Label and the Distance This Paper Holds

The dependency frame carries a history of political use extending well beyond the assistance sector, and a paper supplying a criterion adjacent to that frame incurs an obligation to mark its distance explicitly. This subsection states the function the label performs, the mechanism by which a diagnostic vocabulary is drawn into that function, and the provisions this paper adopts against it.

The label performs a licensing function: a recipient described as dependent is thereby described as a party whose assistance may be reduced on grounds internal to the recipient’s own conduct. Katz traced the operation of that function through the American poverty discourse, in which the vocabulary of the undeserving poor organised eligibility across successive policy régimes (Katz, 1989). Van Oorschot established the criteria by which publics sort claimants, finding control over need, attitude toward help, reciprocity, identity, and level of need to govern popular judgements of deservingness across European populations [empirically tested: comparative survey research; the criteria replicate across countries] (van Oorschot, 2000). Harvey and Lind observed the same function inside the relief sector, where the anticipation of dependency justified the supply of assistance below assessed need (Harvey and Lind, 2005).

The mechanism drawing a diagnostic vocabulary into that function is straightforward.

A classification separating arrangements into two kinds, one of which carries an adverse name, becomes available as a ground for allocation as soon as the classification is applied to recipients in place of arrangements. The present paper adopts three provisions against that transition. The unit of classification is an arrangement, established in Definition 4.1 as a transfer whose recipient is a relation, so that the criterion returns a property of a design. The classification is graded, established in Remark 4.4, so that its output resists conversion into a binary eligibility mark. And the ethical constraint of Section 8 states the firewall in explicit terms, under which the classification of an arrangement holds independent of any allocation decision concerning the parties within it.

2.5 The Reformulation Adopted

The results of Sections 2.2 through 2.4 support a single reformulation, stated here and held throughout the remainder of the paper.

Claim 2.1. (Standing of the general dependency claim). The proposition that transfers as a class produce durable reductions in recipients’ own effort stands without support in the assembled randomised and review evidence. Treatment of that proposition as an established regularity is thereby foreclosed.

Constraint 2.2. (Dependency as a property of design). Dependency is treated throughout this paper as a property of the design and incentive structure of a support arrangement, ascertained arrangement by arrangement. Every use of the term carries that sense, and the criterion of Section 4 supplies the operation by which the property is ascertained.

The reformulation preserves the phenomenon that the dependency frame registers, since arrangements whose withdrawal collapses the relations they sustain exist and are common, while removing the inference from that phenomenon to a general tendency of transfers. It equally preserves the paper’s obligation under Section 2.4, since a property of a design supplies grounds for redesign, and grounds for reduction of assistance to the parties an arrangement covers require an argument the criterion withholds.

2.6 Summary of Results

This section established the following. The dependency claim circulates in a strong form concerning transfers as a class, a moderate form concerning displacement of local provision, and a weak form concerning particular designs. The humanitarian relief review found reliance on relief standing below the assumed level and found the erosion of initiative without demonstration, while identifying design features as the operative variables. The randomised cash transfer evidence found effects on labour supply clustered near zero and effects on expenditure toward alcohol and tobacco clustered at or below zero, with variation across designs supplying the positive finding. The dependency label performs a licensing function documented in the poverty and deservingness literatures, against which this paper adopts three provisions: classification of arrangements in place of recipients, gradedness of output, and the explicit firewall of Section 8. The general claim was accordingly foreclosed as an established regularity (Claim 2.1), and dependency was fixed as a property of design ascertained arrangement by arrangement (Constraint 2.2). Section 3 turns to the accounts already holding the distinction that such an ascertainment presupposes.

3 Prior Holders of the Extractive and Generative Distinction

The role of this section within the paper is to establish the extent of the territory already occupied, so that the contribution stated in Sections 4 through 7 stands at its actual width. Its objective is to exhibit five independent bodies of work holding the distinction between support that maintains a recipient’s position and support that builds a recipient’s capacity, and to identify the single element the five hold in common as an open question. The section proceeds in seven parts. Sections 3.1 through 3.5 treat the five holders in turn: the intergroup helping literature in social psychology, the doctrine of self-help effort in Japanese development cooperation, the protectionand-empowerment pair in human security, the social investment paradigm in welfare policy together with the critique it has attracted, and the account of extractive and generative ownership in the enterprise-design literature. Section 3.6 states the form the five share and the operation each leaves open. Section 3.7 summarises the results. The method is exposition followed by comparison, with each holder labelled by its type of evidence, since the five stand at widely separated points on that scale and the assistance literatures cite them in a single register.

3.1 Autonomy-Oriented and Dependency-Oriented Help in the Intergroup Helping Literature

Nadler’s model of intergroup helping relations supplies the distinction in its most fully developed form, together with experimental support. The model separates dependencyoriented help, which supplies the recipient with a complete solution to the presented problem, from autonomy-oriented help, which supplies the recipient with the tools by which a solution is reached (Nadler, 2002). The central claim of the model concerns the function of the first form: help supplying complete solutions operates as an assertion of the giver’s standing, and a giver whose status advantage is under challenge has reason to prefer it. Nadler and Halabi tested the receptivity side of this account, finding that members of a low-status group identifying strongly with that group received assumed dependency-oriented help from a high-status group with reduced willingness, and that the effect strengthened where the status relation was perceived as unstable and its basis as illegitimate [empirically tested: laboratory experiments on intergroup helping; the receptivity effects replicate, and the field generality of the status-maintenance mechanism remains under study] (Nadler and Halabi, 2006).

Two properties of the model bear on the present paper. Its unit is the helping act together with the status relation in which the act occurs, so the model already locates kind in the structure of the arrangement in place of the magnitude of the transfer. And its evidence is experimental, which places it at the strongest point on the evidence scale among the five holders treated here. What the model supplies is a mechanism and a validated pair of categories. Assignment of a described arrangement to one of the two proceeds in the experimental setting by construction, since the experimenter builds the help condition, and the model leaves the assignment of an arrangement encountered in the field to the analyst’s reading of it.

3.2 The Doctrine of Self-Help Effort in Japanese Development Cooperation

Japanese development cooperation has carried the distinction in doctrinal form for more than three decades. The 1992 Official Development Assistance Charter established support for the self-help effort (自助努力) of partner countries as a governing principle, directing assistance toward 自立的 development, and the 2003 revision retained the principle while placing human security among the basic policies [doctrine] (Government of Japan, 1992). The Development Cooperation Charter of 2015 carried the principle forward, and the 2023 revision restated it with emphasis on investment in people (人への投資) [doctrine] (Ministry of Foreign Affairs of Japan, 2015). The doctrine reaches beyond declaration into evaluation practice: the Japanese rendering of the sustainability criterion in project evaluation is 自立発展性, which names the capacity of an arrangement to develop on its own footing after cooperation ends.

The Japanese formulation stands closest to the operation this paper supplies, since 自立発展性 is assessed at and after project completion and therefore asks a question about the standing of an arrangement once the transfer stops. Two features hold it short of a criterion. The assessment is conducted as one dimension of ex-post project evaluation, so it reports on a completed intervention in place of classifying an arrangement in operation. And the doctrine attaches the property to the project, where the present paper attaches it to the relation the project maintains, a difference that Section 5 shows to matter wherever a project ends while the relation it built continues under other supply.

The same doctrine has generated the most developed self-critique available in any of the five literatures, in the domestic Japanese setting in place of the international one. Section 9 takes up that critique, which concerns the collapse of self-reliance support (自立支援) into expenditure reduction wherever 自立 is defined as exit from benefit receipt.

3.3 The Protection and Empowerment Pair in Human Security Doctrine

The Commission on Human Security, under the chairmanship of Ogata Sadako and Amartya Sen, organised its framework around two pillars standing in the same relation as the pair under discussion. Protection denotes the institutional and normative arrangements shielding people from acute threat, established from above; empowerment denotes the development of people’s capacity to act on their own behalf and on behalf of others, developed from below [doctrine] (Commission on Human Security, 2003).

The report treats the two as mutually reinforcing and assigns to empowerment the role of enabling people to become the authors of their own security.

The pair has attained canonical standing in policy discourse, which raises the cost of any claim to have introduced it. The framework has equally attracted a standing methodological objection concerning analytic overbreadth: a concept encompassing threats from armed violence through disease to economic insecurity supplies limited purchase on the ordering of priorities or the design of instruments (Paris, 2001). That objection bears directly on the present paper’s undertaking, since a pair of pillars admitting almost any programme under one heading or the other yields a vocabulary in place of a classification. The response adopted here is narrowness of scope: the criterion of Section 4 applies to a single support arrangement and returns a graded assignment, and the paper claims reach for it at that level alone.

3.4 The Social Investment Paradigm and the Matthew-Effect Critique

The social investment paradigm in welfare policy holds the distinction at the level of an entire policy régime. Its central proposition treats social expenditure directed at capacitation, in the form of early childhood education and care, active labour market policy, training, and work-family reconciliation, as investment yielding future returns, and distinguishes such expenditure from compensatory transfers maintaining income [doctrine] (Giddens, 1998; Esping-Andersen et al., 2002; Morel et al., 2012; Hemerijck, 2013). The paradigm governed European social policy discourse through the 2000s and 2010s and supplies the extractive-and-generative fork in the register closest to public finance.

The critique the paradigm attracted is required reading for the present paper, since it identifies a failure mode the fading test alone would leave unexposed. Capacitating services distribute themselves according to the capacity to take them up, so households already advantaged in information, scheduling, and proximity use them at higher rates, and a régime shifting expenditure from transfers toward services can widen the inequality it addresses [empirically tested: comparative European evidence on childcare and active labour market policy use gradients; the direction of the effect is well supported and its magnitude varies by country and instrument] (Bonoli et al., 2017; Cantillon and Van Lancker, 2017). The finding carries a general lesson: an arrangement passing a capacity-building test at the level of each participating relation may fail at the level of the population from which participants are drawn. Section 10 states this as a boundary on the criterion, and the levels firewall observed throughout this paper follows from it.

3.5 Extractive and Generative Ownership in the Design of Enterprise

Kelly’s account of ownership design supplies the terminology of the present paper almost exactly, and citation of it is accordingly mandatory. The account separates extractive ownership, whose purpose is the maximisation of financial extraction from an enterprise, from generative ownership, whose purpose is the generation of the conditions on which life depends, and it develops the latter through five design patterns: a living purpose, rooted membership, mission-controlled governance, stakeholder finance, and ethical networks [doctrine] (Kelly, 2012). The examples supporting the account run through cooperatives, employee-owned firms, community land trusts, and foundationowned companies.

Two properties place the account within the same family as the present paper and hold it apart from it. Kelly’s unit is the ownership architecture of an enterprise, and the classification proceeds by inspection of that architecture against the five patterns, which yields a typology of designs. The present paper’s unit is a support arrangement, and the classification proceeds by a counterfactual applied to the supported relation, which yields a graded assignment. The distinction the two share is thereby held at different levels and ascertained by different operations, and the terminological overlap is acknowledged here in full.

3.6 The Common Form of the Five Accounts and the Operation They Leave Open

Table 1 sets the five accounts alongside one another. Three elements recur across all of them. Each locates the difference in the design of an arrangement, holding the magnitude of the transfer aside. Each treats the maintaining form as carrying a cost borne by the recipient’s standing or capacity. And each holds the capacitating form as the objective toward which practice should move. The recurrence establishes the concession this paper opens with.

Claim 3.1. (The status of the distinction). The distinction between support maintaining a recipient’s position and support building a recipient’s capacity is independently held in five literatures, in one of which it carries experimental support.

Any claim of novelty for the distinction itself is thereby foreclosed.

The element the five leave open is visible in the final column of Table 1. Assignment proceeds in each case by a procedure available at one point only: the experimenter’s construction of a condition, a rating at project completion, the heading under which a programme is filed, the instrument through which expenditure flows, or the inspection of an ownership architecture. Each procedure serves the purpose for which its own literature developed it, and each requires either a completed intervention or an inspectable design document.

Claim 3.2. (The unsupplied operation). The five prior holders name the distinction and leave open a test applying to a support arrangement while that

Table 1: The five prior holders of the extractive and generative distinction, with the domain in which each holds it, the terms in which each names it, the type of evidence supporting each, and the operation each supplies for assigning a described arrangement.

Holder Domain Terms naming the pair Type of evidence Operation of assignment Nadler Intergroup helping Dependencyoriented and autonomy-oriented help Experimental Construction of the help condition by the experimenter Japanese development cooperation Official development assistance 自助努力 and 自立発展性 Doctrine, with ex-post evaluation practice Qualitative rating at project completion Ogata and Sen Human security Protection and empowerment Doctrine Categorisation of programmes by heading Social investment paradigm Welfare policy Capacitating services and compensatory transfers Doctrine, with tested critique Classification of expenditure by instrument Kelly Ownership design Extractive and generative ownership Advocacy with case material Inspection of architecture against five patterns arrangement is in operation, returning a graded assignment, and taking as its object the relation the arrangement maintains. Supplying that test is the undertaking of Section 4.

One further concession belongs at this point, concerning the framework within which the test is stated in place of the test itself. Evaluation of arrangements at the level of relations, holding the isolated individual aside as the unit, is independently held by relational egalitarianism (Anderson, 1999), by African relational moral theory (Metz, 2022), by Confucian role ethics (Ames, 2011), by the ethics of betweenness in Watsuji (Watsuji, 1937), and by care ethics (Noddings, 1984; Held, 2006). The relational unit is accordingly conceded together with the distinction, and Section 8 states what the present account adds to it, which concerns the conjunction of that unit with an operational criterion and with a constitutive account of worth.

3.7 Summary of Results

This section established the following. Five independent literatures hold the distinction between maintaining and capacitating support: the intergroup helping literature with experimental support, Japanese development cooperation doctrine with an associated evaluation practice, human security doctrine, the social investment paradigm with its tested Matthew-effect critique, and the ownership-design literature with case material (Claim 3.1, Table 1). The Japanese formulation stands closest to an operation, since 自立発展性 is assessed after cooperation ends, and holds short of one by attaching the property to a completed project in place of an operating relation. The Matthew-effect critique supplies a boundary carried through the paper, by which an arrangement passing a capacity test at the level of each participating relation may fail at the level of the population from which participants are drawn. The five leave open a test applicable to an arrangement in operation, returning a graded assignment, and taking the supported relation as its object (Claim 3.2). The relational unit of evaluation is conceded alongside the distinction. Section 4 supplies the test, and Section 5 identifies the features of an arrangement that settle its result.

4 The Fading Test and Its Provenance in Scaffolding

The role of this section within the paper is to supply the operation whose absence Sections 2 and 3 establish: a test assigning a given support arrangement to one of two kinds. Its objectives are three. The first is to define support and the two kinds in terms admitting application to a described arrangement. The second is to fix the provenance of the test and the boundary within which the transfer from that provenance holds.

The third is to separate the test from three measurements with which the assistance literatures presently conflate it. The section proceeds in five parts. Section 4.1 states the definitions and the test. Section 4.2 traces the test to the study of instructional scaffolding and states what the transfer carries and what it leaves behind. Section 4.3 separates the test from measurement of outcome, from measurement of duration, and from measurement of the recipient’s expressed satisfaction. Section 4.4 states the companion criterion governing the destination of value generated inside a supported relation. Section 4.5 summarises the results. The method throughout is definitional, with each borrowed component labelled by its type of evidence and carrying the boundary stated at its source.

4.1 Definition of Support and of the Two Kinds

Application of a criterion presupposes a specification of the object to which the criterion applies. The specification adopted here places the recipient of a transfer inside a relation whose operation that transfer maintains.

Definition 4.1. (Support). Support is a transfer of resource, instruction, protection, attention, or authority from a supplying party into a relation sustained by a supported party, made with the effect of holding that relation in operation.

Three features of this definition carry weight through the remainder of the paper.

The recipient of the transfer is a relation: a household’s relation to its own subsistence, a ministry’s relation to the population it administers, a clinic’s relation to the patients on its list, a researcher’s relation to a line of inquiry. Questions about the kind of a support arrangement thereby become questions about the standing of a relation, in place of questions about the character of a person. The definition further leaves the supplying party’s purpose out of account, which Section 5 establishes as inert with respect to kind. And the definition covers transfers of authority and attention alongside transfers of resource, so that arrangements carrying small budgets and large decision rights fall within its scope.

Definition 4.2. (Extractive and generative support). Support is extractive where the operation of the supported relation rests on continuation of the transfer, and generative where the transfer builds within the supported relation the capacity by which that relation holds itself in operation.

Definition 4.2. states a property. The operation by which that property is ascertained follows.

Criterion 4.3. (Fading test). A support is generative where its withdrawal leaves the supported relation intact, and extractive where its withdrawal collapses the relation it sustained.

The test carries three features worth stating explicitly. It is counterfactual, so its application to an arrangement in operation proceeds by staged partial withdrawal or by comparison with a matched arrangement already withdrawn. It is directed at the relation, so the welfare of the supported party enters the assessment through the standing of the relation that welfare depends upon. And it is graded.

Remark 4.4. (Fadeability as a graded property). Applied at a schedule of partial withdrawals, the test returns a fraction: the proportion of the transfer whose removal the supported relation absorbs while holding in operation. Full generativity and full extraction occupy the ends of that scale, and the arrangements encountered in practice occupy its interior.

The sustainability-and-exit literature already performs this operation in the field.

Rogers and Coates, reporting a four-country prospective study of programme exit, found that gains persisted after handover where resources, technical and managerial capacity, motivation, and institutional linkages stood in place before the handover occurred [empirically tested: four-country prospective study; consistent across sites] (Rogers and Coates, 2015). The contribution of Criterion 4.3 lies in converting that finding from an evaluation of completed programmes into a definition of kind applicable to an arrangement in operation.

4.2 The Provenance of the Test in the Study of Instructional Scaffolding

The test stated in Section 4.1 originates in the study of tutoring, and the conditions holding in that setting govern the reach of the transfer into assistance settings. This subsection states the provenance, the evidence supporting it, and the boundary the transfer must carry.

Wood, Bruner and Ross introduced the term scaffolding from observation of tutorial interaction, in which a tutor controls the elements of a task lying beyond the learner’s present capacity so that the learner completes the elements within reach (Wood et al., 1976). Three components compose the process: contingency, by which the tutor’s control varies with the learner’s demonstrated capacity; transfer of responsibility, by which control passes to the learner as capacity forms; and fading, by which the tutor’s contribution declines toward zero while task performance holds. The framework stands in a lineage running from Vygotsky’s zone of proximal development, which locates instruction in the interval between independent performance and assisted performance (Vygotsky, 1978). Fading supplies the operation the assistance literatures lack, since fading is precisely a scheduled withdrawal under which the maintained activity either holds or collapses [empirically tested: instructional dyads; the three components are well attested, with magnitudes varying by domain and by measure].

The instructional setting supplies the operation together with a presupposition that assistance settings hold open.

Constraint 4.5. (Boundary of the transfer from the instructional setting). The instructional setting presupposes a learner on a trajectory toward selfregulation, so the fading of tutorial control tracks a competence already forming.

Assistance settings hold the existence of such a trajectory open as the question at issue. The transfer accordingly carries the operation of staged withdrawal and leaves the presupposition of an underlying trajectory behind, so that the test classifies an arrangement as the arrangement presently stands and leaves its future course open.

Retention of this boundary blocks an inference the deservingness literature makes routinely, by which the finding that an arrangement fails the fading test licenses the conclusion that the arrangement ought to end (van Oorschot, 2000; Katz, 1989). The criterion classifies; Section 8 establishes the constraint under which classification stands apart from allocation, and the boundary stated here supplies the first of the two grounds on which that constraint rests.

4.3 Separation of the Test from Outcome, Duration, and Expressed Satisfaction

Practice assesses support arrangements by three measurements already established in evaluation routine: the magnitude of the outcome achieved, the length of the commitment required, and the satisfaction the recipient reports. The objective of this subsection is to establish that each of the three varies independently of kind, so that the fading test addresses a property the existing measurements leave unaddressed. The method is the exhibition of a case in each pair of directions.

Outcome magnitude separates from kind through the reward literature. Deci, Koestner and Ryan, in a meta-analysis of 128 experiments, found that tangible rewards contingent on task engagement reduced free-choice intrinsic motivation, while positive verbal feedback raised it [empirically tested: meta-analysis, 128 experiments; the undermining effect is robust for expected tangible rewards and contested in scope] (Deci et al., 1999). A reward of this form raises measured task output during the period of its supply and lowers the capacity that produced the output, which is an arrangement scoring well on outcome and failing the fading test. The converse case is supplied by arrangements of modest measured effect whose withdrawal the supported relation absorbs entirely.

Duration separates from kind by inspection of the definition. An arrangement running for decades passes the test where the supported relation would hold on withdrawal and the transfer continues by the supplier’s election, as with maintenance transfers into infrastructure already operated locally. An arrangement running for six months fails the test where the relation it sustained collapses at exit. Length of commitment therefore reports the supplier’s schedule and leaves kind undetermined.

Expressed satisfaction separates from kind through the recipient-reaction literature.

Fisher, Nadler and Whitcher-Alagna established that receipt of aid carries a threat to the recipient’s self-esteem, so that recipient evaluations register that threat alongside the assistance’s usefulness [empirically tested: review of experimental literature on recipient reactions] (Fisher et al., 1982). Bolger, Zuckerman and Kessler, following law graduates and their partners by daily diary through the month preceding the New York State bar examination, found that support the provider reported while the recipient registered none was associated with lower depressed mood, while support the recipient did register carried an emotional cost [empirically tested: daily-diary study of couples under a common stressor, with experimental replication in Bolger and Amarel (2007); the mechanism remains contested] (Bolger et al., 2000; Bolger and Amarel, 2007). Registration of support and benefit from support therefore come apart, and satisfaction measurement supplies weak evidence of kind in both directions.

4.4 The Folding Criterion as the Destination-Side Companion

The fading test addresses the durability of a supported relation at the moment of withdrawal. A second question concerns the destination of the value the relation generates while the arrangement operates, and the criterion answering it stands alongside the fading test as its companion. This subsection states that criterion, establishes the independence of the two, and marks the analytic occasions on which each applies.

Criterion 4.6. (Folding criterion). Generated value folds where it returns to the relation that produced it, and converges where it accrues to an attractor lying outside that relation.

Claim 4.7. (Independence of fadeability and folding). Fadeability and folding vary independently, so an arrangement passing one criterion may fail the other.

An arrangement whose supported relation holds on withdrawal, while the value generated inside that relation accrues to the funder’s reputation, to the implementing organisation’s reported metrics, or to a ranking governing the sector’s resource flows, passes the fading test and fails the folding criterion.

The occasions of application separate accordingly. An analyst asking after the standing of a relation at exit applies the fading test. An analyst asking after the direction in which a relation’s product travels during operation applies the folding criterion. Section 6 shows that the accountability structure of the assistance sector bears directly on the second question, since the party receiving the reported product of an arrangement and the party whose capacity the arrangement addresses occupy distinct positions.

4.5 Summary of Results

This section established the following. Support was defined as a transfer whose recipient is a relation, which places questions of kind at the level of relations and holds them clear of judgements about persons (Definition 4.1). The two kinds were defined by the dependence of the supported relation’s operation on continuation of the transfer (Definition 4.2), and the fading test was stated as the operation ascertaining kind (Criterion 4.3), with fadeability established as a graded property admitting assessment by staged partial withdrawal (Remark 4.4). The provenance of the test was traced to the fading component of instructional scaffolding, and the transfer into assistance settings was bounded by the constraint that the instructional presupposition of a trajectory toward self-regulation remains behind, so that the test classifies present arrangements and leaves their future course open (Constraint 4.5). Outcome magnitude, duration of commitment, and expressed satisfaction were each shown to vary independently of kind.

The folding criterion was stated as the destination-side companion of the fading test (Criterion 4.6), and the two were shown to vary independently (Claim 4.7). Section 5 takes up the determinants settling which kind a given arrangement exhibits.

5 Routing, Visibility, and Assigned Position as Determinants of Kind

The role of this section within the paper is to identify the features of a support arrangement that settle the result the fading test returns, so that the criterion of Section 4 acquires purchase on design in place of purchase on classification alone. Its objectives are two. The first is to establish three features as determinants of kind: the routing of the coupling a transfer creates, the visibility of the transfer to the supported party, and the position the arrangement assigns to that party. The second is to establish three further features as inert with respect to kind: the magnitude of the transfer, its content, and the supplying party’s intention. The section proceeds in six parts. Sections 5.1 through 5.3 treat the three determinants in turn, each with the evidence supporting it and the level at which that evidence was obtained. Section 5.4 establishes the inertness of the three remaining features. Section 5.5 states the determinants claim together with the observations that would tell against it, and Section 5.6 summarises the results. The method is analytic for the routing result, which follows from Definitions 4.1 and 4.2, and evidential for the visibility and position results, which rest on experimental and diary studies whose level of obtainment is stated in each case.

5.1 The Routing of the Coupling a Transfer Creates

Every transfer establishes a coupling between parties, and the identity of the parties so coupled settles what happens at withdrawal. This subsection defines that property and establishes it as the first determinant.

Definition 5.1. (Routing). The routing of a transfer is the identity of the relation the transfer strengthens. A transfer is routed to the supplier where it strengthens the relation between the supported party and the supplying party, and routed to the supported relation where it strengthens the relation between the supported party and the counterparties on which that party’s own operation runs.

Claim 5.2. (Routing as a determinant). Routing settles the result of the fading test. A transfer routed to the supplier leaves the supported party coupled to a source whose removal is the withdrawal itself, so the arrangement collapses at withdrawal by construction. A transfer routed to the supported relation leaves in place a coupling the withdrawal does not touch.

The claim follows from Definitions 4.1 and 4.2 together with Definition 5.1, and its content is best exhibited by a pair of arrangements identical in every respect apart from routing. A donor supplying a district clinic with medicines through its own procurement chain strengthens the clinic’s relation to that chain, and withdrawal of the donor removes the clinic’s supply. A donor financing the clinic’s own contracting with a domestic wholesaler strengthens the clinic’s relation to that wholesaler, and withdrawal of the donor leaves a contracting relation whose continuation turns on the clinic’s revenue.

The two arrangements carry equal budgets, deliver equal quantities of medicine, and register equal outcomes during operation, and they separate at withdrawal.

The same structure organises a documented pattern in assistance to higher education, where support flowing through scholarships tenable at institutions in the supplying country strengthens the recipient’s relation to those institutions, and support flowing to the capacity of institutions in the recipient country strengthens the relation between those institutions and the students they serve (Chankseliani, 2022). Routing accordingly supplies the sense in which arrangements described in identical terms of purpose and volume separate into kinds.

5.2 The Visibility of the Transfer to the Supported Party

The degree to which a transfer registers in the supported party’s own account of the situation bears on the capacity the transfer is meant to build, and the evidence establishing this was obtained in interpersonal settings. This subsection reports that evidence, states the mechanism under dispute, and marks the level boundary governing its use in the present paper.

Bolger, Zuckerman and Kessler followed law graduates and their partners by daily diary through the month preceding the New York State bar examination, separating support reported by the provider from support registered by the recipient. Support the provider reported while the recipient registered none was associated with lower depressed mood in the recipient; support the recipient registered carried an emotional cost [empirically tested: daily-diary study of couples under a common stressor; the pattern replicated under experimental manipulation in Bolger and Amarel (2007), and the evidence is dyadic throughout] (Bolger et al., 2000; Bolger and Amarel, 2007). The mechanism remains under dispute, with candidate accounts running through the threat that registered receipt poses to the recipient’s sense of competence, through the attentional cost of the interaction that registration requires, and through the inference the recipient draws about the provider’s assessment of the recipient’s capacity.

The reward literature supplies a convergent result at the level of task motivation.

Deci, Koestner and Ryan found tangible rewards contingent on engagement to reduce free-choice intrinsic motivation, with verbal feedback carrying the contrary effect (Deci et al., 1999). The common element across the two literatures is the salience of the external supply to the party whose capacity is at issue.

Claim 5.3. (Visibility as a determinant at the interpersonal level). At the level of interpersonal support, the salience of a transfer to the supported party bears on the capacity that party sustains, with reduced salience associated with improved adjustment. The evidence supporting this claim was obtained in dyadic settings, and its extension to organisational and international arrangements stands as a conjecture requiring its own test.

The level boundary stated in Claim 5.3 is observed throughout the paper. Section 10 states the observations that would establish or defeat the organisational extension, and Section 9 draws design implications at the interpersonal level alone.

5.3 The Position the Arrangement Assigns to the Supported Party

Support arrangements distribute the roles of problem definition, method selection, and execution, and the distribution assigns the supported party a position within the arrangement. This subsection establishes that position as the third determinant, drawing on the experimental evidence reported in Section 3.1 and on the evidence concerning the effects of occupying a giving position.

Nadler’s distinction between help supplying complete solutions and help supplying tools is a distinction of assigned position, since the first assigns the recipient the position of a party awaiting a solution and the second assigns the position of a party executing one (Nadler, 2002; Nadler and Halabi, 2006). The operational content of the distinction lies in four allocations: the party defining the problem, the party selecting the method, the party holding the decision at each execution step, and the party accounting for the result. An arrangement allocating all four to the supplier assigns the supported party a receiving position throughout.

Riessman’s helper-therapy principle establishes the far end of the same scale. Parties occupying a giving position within an arrangement register gains attributable to that position, so that arrangements placing recipients in positions of supply toward others operate on the recipients’ own capacity [empirically tested: clinical and programme observation, with subsequent study in peer-support settings; the direction of the effect is consistently reported and its magnitude varies by setting] (Riessman, 1965).

Claim 5.4. (Assigned position as a determinant). The position an arrangement assigns to the supported party, ascertained by the allocation of problem definition, method selection, execution decisions, and accounting, bears on the capacity that party sustains and thereby on the result of the fading test.

5.4 The Inertness of Magnitude, Content, and Intention

Practice discriminates among support arrangements by three features standing outside the list established above, and the objective of this subsection is to establish each as inert with respect to kind. The method is exhibition of arrangements standing at opposite ends of each feature while sharing a kind.

Magnitude stands inert. A transfer of substantial volume routed to the supported relation, as with budget support financing a ministry’s own contracting, leaves that relation in place at withdrawal. A transfer of modest volume routed to the supplier, as with a small grant administered through the supplier’s own staff and systems, collapses at withdrawal. Volume accordingly reports the scale of an arrangement and leaves its kind open.

Content stands inert. Cash transfers admit both routings, since a transfer paid into a household’s own account against the household’s own purposes strengthens that household’s existing exchange relations, while a transfer paid against eligibility conditions the supplier administers and verifies strengthens the household’s relation to the supplier’s administration. Training admits both routings, since instruction yielding a certification recognised across the domestic labour market strengthens the trainee’s relation to that market, while instruction yielding a certification recognised by the supplying organisation alone strengthens the trainee’s relation to that organisation. The medium through which support flows therefore leaves kind open, which places the cash-versus-services debate in welfare policy outside the present criterion’s frame of reference.

Intention stands inert, and this result carries the greatest weight for practice. The fading test is a counterfactual concerning the standing of a relation, and the supplying party’s purposes enter that counterfactual at no point. The experimental evidence reinforces the analytic result, since the status-maintaining function Nadler identifies operates through the structure of the help supplied and stands available to suppliers holding entirely benevolent purposes (Nadler, 2002). Section 6 supplies the organisational mechanism through which a sector holding the capacitating aspiration produces arrangements of the maintaining kind at scale, which is the systematic form of the same result.

Claim 5.5. (Inertness of magnitude, content, and intention). The volume of a transfer, the medium through which it flows, and the purposes of the supplying party stand inert with respect to kind. Assignment of an arrangement to a kind proceeds through routing, visibility, and assigned position alone.

5.5 The Determinants Claim and the Observations Telling Against It

Proposition 5.6. (The determinants of kind). The result the fading test returns for a support arrangement is settled by the routing of the coupling the transfer creates, by the visibility of the transfer to the supported party, and by the position the arrangement assigns to that party.

Proposition 5.6. admits defeat by observation, and the forms such observation would take are stated here and carried into Section 10. A set of arrangements matched on routing, visibility, and assigned position while separating systematically at withdrawal would establish a determinant the proposition omits. A set of arrangements separating on routing while returning matched results at withdrawal would defeat Claim 5.2.

And a demonstration that arrangements identical in the three determinants separate at withdrawal according to the supplier’s stated purposes would defeat Claim 5.5. The proposition holds the three determinants as jointly settling the result, and it offers no ordering among them, which Section 10 records as an outstanding element of the account.

5.6 Summary of Results

This section established the following. Routing was defined as the identity of the relation a transfer strengthens (Definition 5.1) and established analytically as a determinant of kind, with a matched pair of clinic supply arrangements exhibiting the separation it produces (Claim 5.2). Visibility was established as a determinant at the interpersonal level on diary and experimental evidence, with its extension to organisational and international arrangements marked as a conjecture and its mechanism recorded as under dispute (Claim 5.3). Assigned position was established as a determinant through the intergroup helping evidence and the helper-therapy literature, with its operational content fixed by the allocation of problem definition, method selection, execution decisions, and accounting (Claim 5.4). Magnitude, content, and intention were established as inert, which places the cash-versus-services debate outside the criterion’s frame and locates the sector’s production of maintaining arrangements in structure in place of purpose (Claim 5.5). The three determinants were stated as jointly settling the test’s result, together with the observations that would tell against the statement (Proposition 5.6). Section 6 supplies the organisational mechanism accounting for the prevalence of arrangements routed to the supplier.

6 Accountability Structure and the Drift Toward Extractive Arrangement

The role of this section within the paper is to account for a pattern the preceding sections leave standing as a puzzle: a sector holding the capacitating aspiration in its own doctrine produces arrangements of the maintaining kind at scale, and does so while the parties composing it hold the aspiration sincerely. Its objectives are three.

The first is to establish the asymmetry in the direction of accountability that organises the sector. The second is to trace the transmission from that asymmetry to each of the three determinants established in Section 5, so that the explanation operates through the mechanism the criterion identifies in place of operating alongside it. The third is to state the drift result together with the observations that would tell against it. The section proceeds in five parts. Section 6.1 establishes the asymmetry. Section 6.2 traces the transmission to routing, to visibility, and to assigned position. Section 6.3 treats the consequences for organisational learning. Section 6.4 states the result and its falsifiers. Section 6.5 summarises the results. The method is organisational analysis drawing on the accountability literature in the study of non-governmental organisations, with the level at which each finding was obtained stated at its point of use.

6.1 The Direction of Accountability in the Assistance Sector

Accountability relations in the assistance sector run in two directions, and the two are supported by structures of markedly unequal strength. Upward accountability runs from an implementing organisation to the parties supplying its funds, comprising donors, governmental agencies, and boards, and it is carried by disbursement conditions, reporting requirements, audit, and the renewal decision. Downward accountability runs from the implementing organisation to the parties its programmes address, comprising communities, service users, and partner organisations at the point of delivery, and it is carried by consultation practices, grievance procedures, participation mechanisms, and such exit options as the setting affords.

Ebrahim assembled the analysis of these mechanisms, examining the instruments through which each direction operates and the demands each places on organisational attention (Ebrahim, 2003). The finding is an asymmetry of enforcement in place of an asymmetry of intention: the instruments carrying upward accountability attach consequences to performance, in that a failure of reporting or of disbursement compliance bears on the organisation’s continuation, and the instruments carrying downward accountability attach consequences weakly or leave them to the organisation’s own election [empirically tested: case-based organisational analysis of the non-governmental sector; the asymmetry is consistently reported and its magnitude varies with funding structure] (Ebrahim, 2003; Edwards and Hulme, 1996).

Claim 6.1. (The asymmetry of accountability). In the assistance sector the party supplying the funds for an arrangement and the party whose capacity that arrangement addresses occupy distinct positions, and the accountability running toward the first is carried by enforced instruments while the accountability running toward the second is carried by weakly enforced instruments.

6.2 The Transmission from the Asymmetry to the Three Determinants

Claim 6.1. states a feature of organisational structure, and the explanatory work of this section requires the transmission of that feature into the determinants of kind. This subsection traces the transmission along three paths, one for each determinant established in Section 5, and the argument is analytic throughout: each path proceeds from a requirement that upward accountability imposes to the design feature that satisfies the requirement.

The path to routing runs through attribution. An organisation accountable upward is required to demonstrate that the results it reports followed from the funds it received, which is a demand for attributable causation. Attribution is secured most reliably where the organisation controls the chain through which the transfer flows, since a chain the organisation controls generates records the organisation can produce and outcomes the organisation can trace. A transfer routed to the supported relation’s own counterparties, by contrast, generates results distributed across parties the organisation neither controls nor documents. The reporting requirement accordingly favours routing to the supplier, which Claim 5.2 establishes as the routing under which an arrangement collapses at withdrawal by construction.

The path to visibility runs through verification and identification. An organisation accountable upward is required to verify that transfers reached their intended recipients and, under a widespread further requirement, to identify the source of the support at the point of delivery. Verification places the recipient in the position of supplying evidence of receipt, and identification places the origin of the support in the recipient’s field of attention as a condition of the arrangement’s compliance. Both requirements raise the salience of the transfer to the supported party. Claim 5.3 establishes the adverse direction of raised salience at the interpersonal level and marks its extension to organisational arrangements as a conjecture, so the transmission traced here establishes the direction in which the requirements move the determinant, and the consequence of that movement for organisational arrangements stands as the conjecture Claim 5.3 records.

The path to assigned position runs through the reporting format. Results frameworks specify indicators in advance of implementation, which locates problem definition and method selection with the parties designing the framework. The supported party’s role within such a framework is the supply of the data the indicators require and the performance of the activities the framework enumerates, which is the allocation Claim 5.4 establishes as bearing on kind, in the direction the intergroup helping evidence marks as adverse.

Proposition 6.2. (Transmission of the accountability asymmetry to the determinants of kind). The requirements through which upward accountability operates favour transfers routed to the supplier, raise the salience of transfers to the supported party, and assign that party a receiving position. Each of the three determinants is thereby moved in the direction Section 5 establishes as extractive, with the movement of the visibility determinant carrying the consequence Claim 5.3 records as a conjecture at the organisational level.

Proposition 6.2. supplies the explanation the section undertook. The production of maintaining arrangements by a sector holding the capacitating aspiration requires no failure of purpose on any party’s part, since the requirements generating the three effects are the ordinary requirements of financial responsibility and are held by parties acting on entirely defensible grounds. Claim 5.5 established the inertness of intention analytically; Proposition 6.2 exhibits the structural form of the same result.

6.3 The Consequences for Organisational Learning

The asymmetry established in Section 6.1 bears on a second question beyond the kind of arrangements produced, concerning the capacity of the sector to detect the kind of arrangements it produces. Ebrahim developed this consequence under the heading of accountability myopia, arguing that the concentration of accountability effort in short-term, quantifiable, upward-directed reporting displaces the longer-horizon and less legible information on which organisational learning depends (Ebrahim, 2005).

The consequence for the present paper is specific. The fading test requires observation of a supported relation through and beyond withdrawal, which is an observation falling outside the reporting horizon that upward accountability establishes, since the reporting relationship ordinarily terminates with the funding relationship. The information the criterion requires is thereby generated at the point at which the structure generating information ceases to operate. The sustainability literature reviewed in Section 4.1 exists because a small number of studies were funded to make exactly that observation (Rogers and Coates, 2015), and its small volume relative to the volume of implementation reporting is a consequence of the structure described here.

The folding criterion of Section 4.4 applies to the same structure and returns a definite result. The value generated within programme relations is compiled, reported, and accrues as evidence of performance to the parties standing in the upward relation, where it bears on renewal, reputation, and the terms of subsequent funding. The relation producing that value receives the report as an object of extraction in place of a return, which is convergence in the sense Criterion 4.6 defines.

6.4 The Drift Result and the Observations Telling Against It

Claim 6.3. (Structural drift toward extractive arrangement). A sector in which accountability is enforced upward and unenforced downward produces support arrangements of the extractive kind at rates exceeding those its own doctrine would generate, and produces them through the ordinary operation of its financial and reporting requirements in place of through any failure of purpose among the parties composing it.

Claim 6.3. admits defeat by observation, and three forms of observation would supply it. A comparison of organisations funded through structures differing in the strength of downward accountability, holding sector and setting constant, should show the kind of arrangements produced varying with that strength; matched distributions would tell against the claim. A comparison of arrangements funded under core or unearmarked grants with arrangements funded under project grants carrying attribution requirements should show routing varying accordingly. And a natural experiment in which reporting requirements were altered for one class of grants should show the corresponding change in the three determinants. The claim is stated at the level of the funding structure, and its transfer to the individual organisation requires the further argument that Section 10 records as outstanding.

6.5 Summary of Results

This section established the following. Accountability in the assistance sector is asymmetric in enforcement, running toward the parties supplying funds through instruments attaching consequences and toward the parties programmes address through instruments leaving consequences to the organisation’s election (Claim 6.1). The asymmetry transmits to all three determinants of kind: the attribution requirement favours routing to the supplier, the verification and identification requirements raise the salience of the transfer, and the results framework assigns the supported party a receiving position (Proposition 6.2). The accountability structure further displaces the observation the fading test requires, since that observation falls beyond the horizon at which reporting terminates, and the value generated within programme relations converges on the parties standing in the upward relation, which is a failure of the folding criterion. The drift result was stated together with three forms of observation that would tell against it, and its level of statement was fixed at the funding structure (Claim 6.3). Section 7 turns to the forms in which return from generative arrangements arrives.

7 Generative Reciprocity and the Cross-Scale Axis of Return

The role of this section within the paper is to treat the forms in which value generated within a supported relation returns, and to fix the extent of the paper’s claim on that subject at its defensible width. Its objectives are three. The first is to define the axes along which return travels and to separate them cleanly, since the assistance literatures treat return as a single undifferentiated phenomenon. The second is to concede two of the three axes to the existing theory of reciprocity, which covers them with formal models and empirical support this paper adds nothing to. The third is to state the cross-scale axis as the one on which an addition is claimed, and to state that addition with its limits attached. The section proceeds in five parts. Section 7.1 defines generative reciprocity and separates the three axes. Section 7.2 concedes the cross-time and cross-subject axes. Section 7.3 states the cross-scale axis with its instances. Section 7.4 states the limits of the claim and the adjacent literatures with which the claim competes. Section 7.5 summarises the results. The method is definition followed by comparison against the formal literature on reciprocity, with the status of each borrowed result labelled at its point of use.

7.1 Generative Reciprocity and the Three Axes of Return

Reciprocity in its ordinary sense concerns the discharge of an obligation: a party having received returns something to the party having given, and the transaction closes.

Support arrangements of the kind treated in this paper generate returns of a different character, and the difference is definitional in place of being a matter of degree.

Definition 7.1. (Generative reciprocity). Return is generative where what returns increases the capacity of some relation to generate further, and closing where what returns discharges an obligation established by the original transfer.

Generative return leaves the field of relations with greater capacity than it held before the transfer, and closing return restores the field to the position it held before.

Definition 7.1. separates a phenomenon from a settlement, and the separation opens the question this section addresses: where does generative return arrive? Three axes answer the question, and their separation is the analytic contribution of this subsection. The axis of time concerns the interval between transfer and return, which may extend across years or across generations, so that the return arrives at a point at which the original transfer stands beyond the parties’ accounting.

The axis of subject concerns the identity of the party at which the return arrives, which may be a third party standing outside the original transfer, as when a supported party subsequently supports others.

The axis of scale concerns the level of organisation at which the return arrives, which may differ from the level at which the transfer was made, as when a transfer made to a person returns as a practice held by a group, an arrangement held by an institution, or a distinction available to a field.

The three axes are independent, so a single return may travel along all of them or along one alone.

7.2 The Cross-Time and Cross-Subject Axes as Already Held

The first two axes are covered by an existing body of theory carrying formal models and empirical support, and the concession is stated here before any claim is entered.

Nowak and Sigmund developed the theory of indirect reciprocity, in which cooperation is sustained among parties who may never meet again through the mechanism of reputation: a party’s standing rises through acts of helping and governs the help that party subsequently receives from third parties [formally modelled] (Nowak and Sigmund, 1998, 2005). Boyd and Richerson established the corresponding result for the upstream form, in which a party who has received passes assistance onward to others in place of returning it to the giver [formally modelled] (Boyd and Richerson, 1989).

The two mechanisms cover the axis of subject directly and the axis of time in the same movement, since both operate across intervals at which the original transfer has passed out of the parties’ immediate accounting.

The gift-theoretic literature holds the same territory in a different register, treating the obligation to give, to receive, and to reciprocate as the structure organising exchange in societies whose institutions of settlement differ from those of contract (Mauss, 1925), and the subsequent literature extended the analysis to institutionalised giving among strangers (Titmuss, 1970; Godbout and Caillé, 1992).

Claim 7.2. (The conceded axes). Return arriving after an extended interval and return arriving at a party outside the original transfer are covered by the theory of indirect and upstream reciprocity, which supplies formal models and empirical support for both. This paper claims no addition on either axis.

7.3 The Cross-Scale Axis

The axis of scale is distinguished from the two conceded axes by the identity of what returns. Indirect and upstream reciprocity concern the movement of assistance among parties standing at a common level of organisation, and the return they describe is an act of the same type as the original transfer. Return along the axis of scale is of a different type from the transfer that produced it, and it arrives at a level of organisation the original transfer did not address.

Claim 7.3. (The cross-scale axis of return). Value generated within a supported relation returns in some cases at a level of organisation above the level at which the transfer was made, and it returns there as an object of a different type from the transfer: a transfer to a person returning as a practice adopted by a group, an arrangement adopted by an institution, or a distinction available to a field.

Three instances exhibit the pattern. Support to an individual practitioner whose method is subsequently adopted as the standard procedure of a department returns at the level of the department, and what returns is a procedure in place of a transfer. Support to a single trial site whose protocol is subsequently written into national guidance returns at the level of the health system, and what returns is a guideline. Support to an unaffiliated inquirer whose distinction enters the working vocabulary of a field returns at the level of the field, and what returns is a distinction available to every subsequent inquirer at no cost.

The relation between this axis and the criterion of Section 4 is worth stating explicitly, since the two operate in opposite directions and their conflation would be natural. The fading test asks whether a supported relation holds at withdrawal, which concerns the durability of the relation itself. The cross-scale axis concerns the destination of what the relation generates while it operates, which places the axis on the side of the folding criterion of Section 4.4. A return arriving at a higher level of organisation folds where the field so altered includes the relation that produced the alteration, and converges where the field so altered stands closed to that relation.

7.4 The Limits of the Cross-Scale Claim

The claim entered in Section 7.3 is an observation, and this subsection states the four respects in which it stands short of a theory, together with the adjacent literatures against which any development of it would be assessed.

The claim carries no formal treatment. Indirect reciprocity is supported by models specifying the conditions under which the mechanism sustains cooperation against invasion by parties who decline to help, and the cross-scale axis has no counterpart to those conditions. A formal treatment would specify the conditions under which a return of one type at one level arises from a transfer of another type at another, and this paper supplies none.

The claim carries no frequency. The instances of Section 7.3 establish that the pattern occurs, and the proportion of supported relations generating such returns stands unestimated. Design proceeding on an expectation of cross-scale return would accordingly proceed on an expectation the present account leaves unquantified.

The claim carries no selection mechanism. The instances describe adoption at a higher level, and the process selecting which practices, arrangements, and distinctions are adopted stands outside the account. The absence matters, since the selection process is where the interesting variation is likely to lie.

The claim competes with two established literatures whose coverage overlaps it.

The diffusion literature treats the adoption of practices across populations of adopters and supplies an account of the attributes and channels governing the rate of adoption (Rogers, 2003). The institutional literature treats the mechanisms through which organisations within a field converge on common forms, and its account of mimetic and coercive processes covers a substantial portion of the instances Section 7.3 describes (DiMaggio and Powell, 1983). A development of the cross-scale axis would be required to establish what it adds to those accounts, and the candidate addition lies in the relation between the level at which support was supplied and the level at which the return arrives, which the diffusion and institutional literatures treat as a question about adopters in place of a question about supply.

Constraint 7.4. (The status of the cross-scale claim). The cross-scale axis is entered as an observation. It carries no formal treatment, no estimate of frequency, and no account of selection, and no argument elsewhere in this paper rests upon it.

7.5 Summary of Results

This section established the following. Generative return was defined by its effect on the capacity of relations to generate further, and separated from closing return, which discharges an obligation and restores the prior position (Definition 7.1). Three independent axes of return were separated: time, subject, and scale. The first two were conceded to the theory of indirect and upstream reciprocity, which supplies formal models covering both (Claim 7.2). The cross-scale axis was stated as the axis on which this paper enters an addition, distinguished by the arrival of a return of a different type at a level of organisation above that of the transfer, and exhibited through three instances (Claim 7.3). The axis was located on the folding side of the paper’s apparatus in place of the fading side. Four limits were stated: the absence of a formal treatment, of a frequency estimate, and of a selection mechanism, together with the overlap against the diffusion and institutional literatures, and the claim was fixed as an observation on which nothing else in the paper rests (Constraint 7.4). Section 9 turns to the implications of the preceding results for the design of assistance.

8 The Ethical Constraint on the Use of the Criterion

The role of this section within the paper is to establish the constraint under which the criterion of Section 4 may be employed, since a classification separating support arrangements into a favoured and a disfavoured kind stands adjacent to decisions about who receives support and in what measure. Its objectives are four. The first is to ground the worth of the parties within a supported relation in advance of any criterion and independently of it. The second is to state the argument from systemic generative capacity and to decline it as a ground. The third is to establish the diagnostic status of the criterion against the objection that a self-maintenance test cannot yield an evaluative conclusion. The fourth is to state the firewalls holding the criterion clear of allocation, together with the limit case that tests them. The section proceeds in eight parts. Section 8.1 grounds worth. Section 8.2 establishes that receipt stands clear of any requirement of return. Section 8.3 states and declines the systemic justification. Section 8.4 treats the derivation objection. Section 8.5 states the lexical firewall. Section 8.6 treats the limit case of support admitting no withdrawal and names the two errors available within it. Section 8.7 states the deservingness firewall. Section 8.8 summarises the results.

The method is philosophical argument drawing on the value-theoretic and dependency literatures, with the structural devices employed identified by name at each point of use.

8.1 The Ground of Worth, Stated in Advance of the Criterion

Any account locating the unit of evaluation in relations faces an objection stated most sharply in the African relational literature under the heading of the focus problem:

grounding a person’s worth in that person’s relations appears to value the person for the sake of the relations, which fails to value the person for her own sake. The objection presses with particular force on the present paper, since Definition 4.1 makes a relation the recipient of every transfer. The device answering it is due to Korsgaard.

Korsgaard separates two distinctions ordinarily run together. The intrinsic and extrinsic distinction concerns the source of a thing’s value, according to whether that value resides in the thing itself or is conferred from outside it. The final and instrumental distinction concerns the way a thing is valued, according to whether it is valued as an end or valued for the sake of something further. The two distinctions are orthogonal, so a thing may hold value whose source lies outside it while being valued as an end (Korsgaard, 1983).

Claim 8.1. (Final value with a relational source). A participant within a supported relation holds final value whose source lies in relations. The relational source of that value leaves its finality intact, so the participant is valued as an end at every point at which the relational source is acknowledged.

Claim 8.1. dissolves the dilemma the focus problem constructs. A framework may hold that a person’s standing arises within relations and hold equally that the person is never valued for the sake of what the relations produce. The load the claim carries through the remainder of the section is the following: worth is established here, in advance of Section 4’s criterion and independently of any result that criterion returns, so that every subsequent argument about arrangements operates on a field in which worth is already fixed.

8.2 Receipt Standing Clear of Any Requirement of Return

The dependency literature carries an implicit expectation that receipt establishes an obligation to return, and the disability and dependency literatures have addressed that expectation directly. This subsection reports their result and states its consequence for the present paper.

Kittay’s account of dependency work establishes a form of reciprocity that runs through third parties in place of running back to the giver. Under the principle of doulia, the party who cares for a dependent is owed support by the wider society, so obligations nest: the cared-for party owes the carer nothing, and the society owes the carer the conditions of the caring (Kittay, 1999). Nussbaum establishes the corresponding result against contractarian foundations, showing that a theory deriving principles from an agreement among parties roughly equal in power and capacity excludes those whose dependency is permanent (Nussbaum, 2006). Fineman establishes the third element, treating vulnerability as a universal condition of embodied life and resilience as a property conferred by institutions in place of a property possessed by individuals (Fineman, 2008). MacIntyre reaches a convergent position from the virtues, treating acknowledged dependence as a standing feature of human life (MacIntyre, 1999).

Claim 8.2. (Receipt without return). Receipt of support establishes no obligation of return running to the supplying party. The value a supported relation generates and the entitlement of its participants to support stand as separate questions, and results concerning the first leave the second untouched.

Fineman’s result carries a further consequence bearing directly on the interpretation of the fading test. The capacity by which a relation holds itself in operation after withdrawal is itself produced relationally, through the institutions, markets, networks, and administrations available in a given setting. An arrangement returning an extractive assignment therefore reports on the relational infrastructure available at that place as much as it reports on the arrangement, and every reading of such a result as a finding about the parties within the relation misassigns the finding. Section 10 records this as a limit on the criterion’s interpretation.

8.3 The Argument from Systemic Generative Capacity, Stated and Declined

An argument available within the present framework would ground the support of parties whose relations return little on the contribution such support makes to the generative capacity of the system as a whole. This subsection states the argument in its strongest form, identifies its structure, and declines it as a ground while retaining it as an observation.

The argument runs as follows. Relations are coupled, so the maintenance of any relation contributes to the conditions under which other relations operate. A society sustaining the relations of its most severely constrained members thereby sustains the trust, the institutional habits, and the sense of standing on which the generativity of the whole depends. Support extended without regard to individual return therefore returns to the system, and the apparent tension between the criterion and the protection of the most constrained dissolves at the level of the system.

The argument’s structure is the source of the difficulty. It grounds protection in an accounting whose boundary has been widened, and an accounting so grounded delivers whatever result the accounting yields. Where the aggregate contribution of sustaining a set of relations is calculated as negative, the same argument licenses abandonment of those relations, and the licence follows from the argument’s form in place of following from any particular empirical estimate. The argument is thereby utilitarian, with the enlargement of the accounting boundary supplying its appearance of generosity.

Claim 8.3. (The systemic justification declined). The proposition that support extended without regard to individual return sustains the generative capacity of the system as a whole is retained as an observation and declined as a ground.

It carries no justificatory weight at any point in this paper, and no protection asserted here rests on it.

The observation retains interest, since it bears on the design of institutions and on the empirical study of their performance. The bar established by Claim 8.3 concerns the direction of support the observation may be asked to supply, and the direction is one way: the observation may be cited in explanation of an institutional pattern, and it stands unavailable as a premise in any argument concluding to an entitlement or to its withdrawal.

8.4 The Derivation Objection and the Diagnostic Status of the Criterion

An objection standing against every framework of the present kind holds that a criterion of self-maintenance cannot yield an evaluative conclusion, since the transition from a description of what sustains itself to a claim about what is good commits the error Moore identified (Moore, 1903). This subsection states the objection precisely, identifies the two available responses, and states the response adopted.

The objection applies to frameworks deriving worth from a self-maintenance property.

Jonas supplies the principal attempt at such a derivation, treating the self-assertion of metabolic life against its dissolution as the origin of value and contesting Moore’s stricture directly (Jonas, 1966). The attempt is instructive and remains contested, and a paper relying on it would inherit the whole of that controversy.

The response adopted here proceeds by the other route, which is available because Section 8.1 grounded worth in advance.

Claim 8.4. (Diagnostic status of the criterion). The fading test classifies support arrangements and confers no worth on any party. Worth is grounded in

Claim 8.1. , prior to the criterion and independently of it, so the derivation objection finds no derivation in this paper to which it applies.

Claim 8.4. fixes a rule of composition observed throughout the paper. Any sentence in which the criterion performs evaluative work concerning persons is an error to be corrected in place of a position to be defended, and the sentences of Sections 6 and 9 concerning the redesign of arrangements are to be read as directed at arrangements throughout.

8.5 The Lexical Firewall

Frameworks evaluating parts by their contribution to a whole attract a standing objection developed against holist environmental ethics, under which such frameworks license the sacrifice of individuals whose contribution is calculated as adverse (Regan, 1983). The structural device answering the objection is lexical priority, employed in the environmental literature to place the protection of individuals ahead of any systemic accounting in a strict order that admits no trade (Callicott, 1989). This subsection states the device in the form it takes here.

Constraint 8.5. (The lexical firewall). Protection of dignity holds prior to every generative accounting, and remains in force at any value of the return that accounting yields.

Constraint 8.5 is exceptionless, and its exceptionlessness is the whole of its content.

A priority admitting exceptions at extreme values is a weighting, and a weighting is precisely what Regan’s objection targets. The constraint accordingly holds at every value the accounting returns, including values a reader might regard as decisive, and no formulation appearing elsewhere in this paper is to be read as qualifying it.

The constraint operates alongside a second requirement drawn from Margalit, under which an institution humiliating those it serves stands condemned as indecent whatever the adequacy of its provision (Margalit, 1996). The requirement bears directly on the determinants of Section 5, since arrangements assigning the supported party a purely receiving position and arrangements making receipt maximally salient are the arrangements through which institutional humiliation ordinarily operates. Decency accordingly constrains design on grounds of its own, and those grounds hold whatever result the fading test returns.

8.6 The Limit Case of Support Admitting No Withdrawal

The criterion meets its hardest case in arrangements whose withdrawal collapses the supported relation under every available design: continuous personal assistance for a person with profound impairment, palliative care, and support to households in settings whose surrounding institutions have themselves collapsed. This subsection states what the criterion returns in such cases, states what follows, and names the two errors available to a reader at this point.

The criterion returns an extractive assignment. The result is a report about the arrangement’s behaviour at withdrawal, which was the whole of what Criterion 4.3 was defined to report. Constraint 8.5 governs what follows from the result, and what follows is nothing bearing on the entitlement of any party within the relation. The support continues, and its continuation rests on Claim 8.1 and Claim 8.2, which were established in advance of the criterion and stand unaffected by its output.

Remark 8.6. (The two errors available at the limit case). Two errors are available to a reader confronting an arrangement whose withdrawal collapses the supported relation under every design. The first reads the extractive assignment as a verdict, treating the classification as a ground for terminating or reducing the arrangement. The second attempts to rescue the arrangement by locating a return arriving elsewhere, in the family, the community, or the society. The two errors share a single premise, under which the standing of an arrangement turns on an accounting of return, and Constraint 8.5 rejects that premise in both directions.

The second error deserves emphasis, since its motive is benevolent and its effect is the graver of the two. A defender of the arrangement who argues that the support does generate a return, arriving at some wider scale, has conceded that an arrangement generating no return would stand differently. The concession converts an unconditional protection into a claim about an empirical magnitude, and empirical magnitudes are open to contrary estimates. The defence thereby surrenders the ground it was constructed to hold.

A distinction of layers organises the practical position and is stated here in place of being left implicit. Protection of the conditions on which a life and its relations continue is supplied unconditionally, and its supply stands independent of every criterion in this paper. The building of capacity is designed against the criterion, since the criterion is precisely an account of which designs build capacity. The amplification of generativity beyond the point of self-maintenance is allocated selectively, since resources are finite and their allocation requires grounds. Unconditional support at the level of standing and finite support at the level of resources are thereby held apart, and the appearance of conflict between them arises from their conflation.

8.7 The Deservingness Firewall

The vocabulary of self-help carries a documented history of allocative use, and a paper supplying a criterion adjacent to that vocabulary incurs an obligation to state its position on eligibility explicitly. This subsection discharges that obligation.

Van Oorschot established the criteria by which publics sort claimants, finding control over need, attitude toward help, reciprocity, identity, and level of need to organise popular judgements of deservingness across European populations [empirically tested:

comparative survey research; the criteria replicate across countries and their relative weight varies] (van Oorschot, 2000). Katz traced the operation of the corresponding vocabulary through successive American policy régimes, in which the classification of the poor into deserving and undeserving organised eligibility (Katz, 1989). The attitude criterion is the one the present paper’s vocabulary is liable to serve, since a doctrine of self-help effort supplies a natural language for assessing a claimant’s disposition.

Constraint 8.7. (The deservingness firewall). No allocation of support proceeds on a prior demonstration of self-help effort or of generativity by any party.

Self-crossing supplies the condition on which support enters a generative cycle, and compassion supplies the ground on which support proceeds in advance of any demonstration of generativity.

The two halves of Constraint 8.7 belong to different registers, and the separation is the point. The first half is a mechanism claim, describing the conditions under which support entering a relation compounds in place of dissipating, and its warrant lies in the evidence of Sections 4 and 5. The second half is an ethical claim, establishing that the supply of support proceeds in advance of any such condition being met. A reading converting the first half into an eligibility test performs exactly the transition the deservingness literature documents, and the second half is stated in the same constraint so that the conversion has a standing obstacle at the point where it would occur.

8.8 Summary of Results

This section established the following. The worth of participants within a supported relation was grounded through Korsgaard’s separation of the source of value from the manner of valuing, so that value with a relational source stands as final value, which answers the focus problem and fixes worth in advance of any criterion (Claim 8.1). Receipt was established as carrying no obligation of return, on the dependency and vulnerability literatures, with the further consequence that an extractive assignment reports on the relational infrastructure available in a setting as much as on the arrangement (Claim 8.2). The argument grounding protection in the generative capacity of the system as a whole was stated in its strongest form, identified as utilitarian in structure, and declined as a ground while retained as an observation carrying no justificatory weight (Claim 8.3).

The derivation objection was answered by the diagnostic status of the criterion, which classifies arrangements and confers worth on no party (Claim 8.4). The lexical firewall was stated in exceptionless form and supplemented by the requirement of decency in institutional design (Constraint 8.5). The limit case of support admitting no withdrawal was treated, the criterion’s output there identified as a report bearing on nothing concerning entitlement, and the two available errors named, together with the premise both share (Remark 8.6). The layers of protection, capacity building, and amplification were separated, with unconditional support at the level of standing held apart from finite support at the level of resources. The deservingness firewall was stated, with its mechanism half and its ethical half distinguished by register (Constraint 8.7). Section 6 turns to the organisational structure through which arrangements of the maintaining kind are produced at scale.

9 Implications for the Design of Assistance

The role of this section within the paper is to draw out what the preceding results imply for the construction of support arrangements, so that the criterion of Section 4 bears on design in place of bearing on assessment alone. Its objectives are four. The first is to state what a design admitting fading requires, drawing on the determinants of Section 5 and on the evidence concerning the conditions under which gains persist after exit.

The second is to establish that participation within an arrangement may precede the capability the arrangement builds, on the strongest evidence available for that ordering. The third is to state the documented failure mode of self-reliance support, drawn from the Japanese literature in which it has been examined most thoroughly. The fourth is to identify the accountability structure as the lever through which the drift established in Section 6 is addressed. The section proceeds in six parts, treating the four objectives in Sections 9.1 through 9.4, fixing the logical form of the implications in Section 9.5, and summarising in Section 9.6. The method combines the analytic results of Sections 4 and 5 with empirical evidence whose type and strength are labelled at each point of use, and every implication stated here is directed at arrangements and stands governed by the constraints of Section 8.

9.1 Design Admitting Fading

An arrangement admitting fading is one whose staged withdrawal leaves the supported relation in operation, and Sections 4 and 5 together specify what such an arrangement requires. This subsection states three requirements and the evidence bearing on each.

The first requirement concerns routing. Claim 5.2 established that a transfer strengthening the relation between the supported party and the supplying party collapses at withdrawal by construction, so an arrangement admitting fading routes its transfer to the counterparties on which the supported party’s own operation runs. The design question at inception is accordingly the identity of the relation the transfer will strengthen, and the question admits an answer at inception, which is the practical gain the criterion supplies.

The second requirement concerns the schedule. A withdrawal conducted at the end of a funding period tests the arrangement once, at the point at which the information the test yields can inform nothing. A schedule of partial withdrawals conducted during operation, in the manner Remark 4.4 describes, yields the same information at points at which redesign remains available. The instructional literature supplies the model, since contingent variation of support against demonstrated capacity is the mechanism through which scaffolding operates (Wood et al., 1976).

The third requirement concerns the conditions standing in place before withdrawal.

Rogers and Coates found that gains persisted after handover where the resources, the technical and managerial capacity, the motivation, and the institutional linkages required by the continuing activity were established before the handover occurred [empirically tested: four-country prospective study of programme exit; consistent across sites] (Rogers and Coates, 2015). The finding converts into a design requirement directly: the elements the relation will require after withdrawal are established during the period of support, and their establishment is the substance of what the support does.

Claim 9.1. (The design implication of routing). An arrangement admitting fading is constructed by routing its transfer to the counterparties of the supported relation, by conducting staged partial withdrawals during operation in place of a single withdrawal at completion, and by establishing the resources, capacities, and linkages the relation will require before the withdrawal occurs.

9.2 Participation Preceding Capability

A design principle in wide use holds that a party is prepared for participation in an activity through training conducted in advance of that participation, and the strongest available evidence on the ordering runs against it. This subsection reports that evidence and states its implication.

Individual Placement and Support is a model of supported employment in which participants are placed in competitive employment at the outset and receive support within the employment thereafter, in contrast with models providing preparatory vocational training in advance of placement. Burns and colleagues tested the two orderings in a randomised trial across six European centres, finding that 54.5 per cent of participants assigned to the place-then-train condition worked at least one day during the follow-up period against 27.6 per cent of those assigned to the train-then-place condition [empirically tested: multi-site randomised controlled trial; the direction of the result is consistent with the wider supported-employment literature and the absolute rates vary with local labour market conditions] (Burns et al., 2007).

Claim 9.2. (Participation preceding capability). Placement of a supported party within the activity, with support supplied inside the activity thereafter, produces higher rates of sustained participation than preparation conducted in advance of placement. The result holds for supported employment on randomised evidence, and its extension to other domains of assistance stands as a conjecture.

The result connects to the determinants of Section 5 through assigned position. A preparatory arrangement assigns the supported party the position of a trainee throughout its duration, and the position of a participant arrives at the end of the arrangement or fails to arrive. A place-then-train arrangement assigns the position of participant at the outset and supplies support inside it, which is the allocation Claim 5.4 identifies as favourable. Braithwaite’s separation of reintegrative from stigmatising responses to wrongdoing exhibits the same fork in the moral domain, where the response either restores the party’s position within the community or fixes the party outside it (Braithwaite, 1989).

9.3 The Documented Failure Mode of Self-Reliance Support

Japanese social policy supplies the most thoroughly examined instance of an assistance system built explicitly on the capacitating aspiration, together with a self-critique identifying the mechanism by which such a system fails and an institutional device constructed against that mechanism. This subsection reports the sequence, since it identifies a failure mode the fading test alone would leave undetected and an existing legal instantiation of the firewall stated in Section 8.7. The account follows Ikeya’s reconstruction of the programme’s introduction and of the debate surrounding it (Ikeya, 2018).

Public assistance administration in Japan long understood self-reliance (自立) as exit from assistance, so that the objective the statute names alongside the guarantee of a minimum standard of living was operationalised as the termination of receipt (Ikeya, 2018).

The report of the specialist committee on the public assistance system in 2004 and the Director-General notice establishing the basic policy for self-reliance support programmes in 2005 replaced that reading with three senses of the objective: economic selfreliance through employment (経済的自立), self-reliance in daily living (日常生活自立), understood as the recovery and maintenance of health and the management of one’s own living, and self-reliance in social life (社会生活自立), understood as the recovery and maintenance of social ties and participation as a member of a local community [doctrine] (Social Security Council, Japan, 2004; Ministry of Health, Labour and Welfare of Japan, 2005). The legislation of 2013 for persons in economic difficulty, in force from 2015, carried the three senses beyond the population receiving public assistance and organised its own provision around accompaniment support (伴走型支援), in which a supporting party remains alongside the supported party through an extended period in place of delivering a discrete intervention [doctrine] (Ministry of Health, Labour and Welfare of Japan, 2015).

The critique advanced from within the administration identifies three points of failure. The programme catalogue issued at introduction was weighted toward employment support, so that the two further senses of the objective stood at risk of becoming ornamental and support to those with low prospects of employment stood at risk of being treated as secondary (Ikeya, 2018). The official chart setting out the flow of self-reliance and employment support terminated in the alteration, suspension, or termination of assistance, so that participation in a programme stood available as a test of continued eligibility (Ikeya, 2018; Social Security Council, Japan, 2004). And administrators framed the objective of contracted employment support in terms of the reduction of public expenditure, a framing carried into the reporting of the programmes, in which the amount of assistance expenditure reduced through participants’ exit or increased earnings appears among the reported outcomes (Ikeya, 2018).

Remark 9.3. (The definitional failure mode). An assistance system optimising toward a defined objective produces the objective as defined. Where selfreliance is defined as the cessation of receipt, the system’s measured success and the reduction of its own expenditure become one quantity, and the capability the objective was introduced to name stands outside what the system optimises.

The device constructed against this mechanism is of direct interest to the present paper, since it separates support from eligibility enforcement by legal ground. Participation in a self-reliance support programme was located under the advisory provision of the Public Assistance Act, which carries no sanction for a recipient who declines, while the directive provision, under which non-compliance may be followed by alteration, suspension, or termination of assistance, was held separate. A recipient whose participation was judged insufficient was thereby moved out of the programme and into the directive route in place of being sanctioned within the programme (Ikeya, 2018).

Ikeya records that the separation held in form and that the boundary between the two routes remained unsettled in practice, since the criteria assigning a recipient from one route to the other were left unspecified.

The sequence carries two lessons into this paper. The first is a caution against the misuse of the criterion advanced here. The fading test asks whether the supported relation holds at withdrawal, and a system measuring its own performance by the volume of withdrawals it achieves satisfies a quantity the test does not name. The test concerns the standing of the relation, and the withdrawal is the instrument through which that standing is ascertained. A system reversing the two, treating withdrawal as the objective and the relation’s standing as the incidental matter, reproduces the failure mode Remark 9.3 describes with the present paper’s vocabulary supplying the terms. The second is that the firewall stated in Constraint 8.7 has an existing legal form, and that the Japanese experience locates its point of weakness precisely: the firewall holds where the criteria assigning a party from the supported route to the enforcement route are themselves specified, and it stands exposed wherever those criteria are left to administrative judgement.

9.4 The Accountability Structure as the Lever

Section 6 established that the three determinants of kind are moved in the extractive direction by the enforcement structure of the assistance sector, which locates the point at which design intervention operates. This subsection states the implication and the three instruments through which it would proceed.

The first instrument concerns the strength of downward accountability. Proposition 6.2 traced the drift to requirements imposed by the enforced upward relation and left unopposed by the weakly enforced downward relation, so instruments attaching consequences to the downward relation address the drift at its source. Mechanisms carrying such consequences include the allocation of decision rights over programme continuation to the parties served, grievance procedures with binding effect, and the placement of funds under the control of those parties.

The second instrument concerns the form of the funding relationship. Attribution requirements attach to project funding earmarked against specified activities and attach weakly to core or unearmarked funding, so the composition of an organisation’s funding bears on the routing its arrangements adopt. Claim 6.3 states the corresponding prediction, and its test is described in Section 6.4.

The third instrument concerns the observation window. Section 6.3 established that the information the fading test requires is generated beyond the horizon at which reporting terminates, so an observation window extending past the end of funding is the condition on which the sector obtains information about the kind of arrangements it produces. The sustainability studies establishing what persists after exit are the existing instances of such observation, and their volume relative to implementation reporting indicates the size of the gap.

Claim 9.4. (The accountability lever). Intervention on the kind of arrangements a sector produces operates through the enforcement structure of its accountability relations, by attaching consequences to the downward relation, by shifting the composition of funding toward forms carrying weaker attribution requirements, and by extending observation beyond the termination of funding.

9.5 The Logical Form of the Implications

The implications stated in this section take a specific logical form, and the form is fixed here so that their standing is unambiguous. This subsection states two constraints governing every implication above.

Constraint 9.5. (Predictions in place of recommendations). The implications of this section are stated as predictions concerning the behaviour of arrangements under specified designs, and each carries the evidence and the level at which that evidence was obtained. Their conversion into recommendations requires a judgement about objectives that this paper supplies at no point.

Constraint 9.6. (Design implications and allocation). Every implication stated in this section is directed at the construction of arrangements. No implication bears on the allocation of support among parties, on the level of support any party receives, or on the eligibility of any party to receive it, and Constraints 8.5 and 8.7 govern the whole of this section.

Constraint 9.6 carries a practical consequence worth stating, since the design implications above are liable to a reading under which arrangements failing the fading test are to be discontinued. The implications concern the construction of arrangements, and the arrangement is the object available for redesign. A finding that an existing arrangement collapses at withdrawal indicates a redesign of routing, schedule, and assigned position, conducted while the support continues, and Section 8.6 treats the cases in which redesign leaves the finding unchanged.

9.6 Summary of Results

This section established the following. A design admitting fading requires routing to the counterparties of the supported relation, staged partial withdrawal during operation, and the establishment of the resources, capacities, and linkages the relation will require in advance of withdrawal (Claim 9.1). Placement within an activity with support supplied inside it produces higher sustained participation than preparation conducted in advance, on randomised multi-site evidence in supported employment, with extension to other domains marked as a conjecture, and the result was connected to assigned position (Claim 9.2). The Japanese self-reliance system supplied the documented failure mode, in which the collapse of three senses of self-reliance into exit from benefit receipt aligns the system’s measured success with the reduction of its own expenditure, and the caution was carried to the present criterion, which concerns the standing of a relation in place of the volume of withdrawals achieved (Remark 9.3). The accountability structure was identified as the lever, operating through consequences attached to the downward relation, the composition of funding, and the extension of the observation window beyond the termination of funding (Claim 9.4). The implications were fixed as predictions carrying their evidence and level (Constraint 9.5), and as directed at arrangements alone (Constraint 9.6). Section 10 states the limits of the account and the observations that would tell against it.

10 The Limits of the Account and the Conditions of Its Falsification

The role of this section within the paper is to state the boundaries of what the preceding sections establish and to specify the observations under which each principal claim would be given up. Its objectives are four. The first is to state the limits on the application of the criterion, which arise from the counterfactual form of the test and from the settings in which withdrawal is unavailable as an operation. The second is to state the transfers among levels of organisation that the paper leaves outstanding. The third is to state the quantities the account leaves unestimated. The fourth is to assemble the falsification conditions attached to the principal claims into a single statement. The section proceeds in five parts, treating the four objectives in Sections 10.1 through 10.4 and summarising in Section 10.5. The method is the systematic exhibition of what the account leaves open, and every limit stated here is a limit on the account in place of a qualification of the results already established.

10.1 Limits on the Application of the Criterion

The fading test is a counterfactual, and its application by the conduct of an actual withdrawal is available in a restricted set of circumstances. This subsection states three limits arising from that form.

The first limit is ethical and is the most consequential in practice. Withdrawal of support from a party for the purpose of ascertaining what follows is barred wherever the party’s standing depends on the support, and Constraints 8.5 and 9.6 bar it within this paper’s own framework. The criterion is accordingly applied through three routes standing at some remove from its definition: comparison with matched arrangements already withdrawn for reasons of their own, staged reduction conducted as part of a planned transition whose completion is itself the objective, and inspection of the three determinants established in Section 5. The third route carries the weight in practice, and it does so at a cost: inspection of routing, visibility, and assigned position yields the criterion’s result through Proposition 5.6, so an error in that proposition propagates into every application conducted by inspection.

The second limit concerns the collapse the definition invokes. Criterion 4.3 separates arrangements by whether the supported relation holds at withdrawal, and relations degrade by degrees in place of holding or failing outright. Remark 4.4 addresses one dimension of the gradation, in the proportion of the transfer whose removal the relation absorbs, and it leaves a second dimension unaddressed, in the degree of function the relation retains under a given reduction. An account specifying the second dimension would state what proportion of a relation’s operation must persist for the relation to count as holding, and the present account supplies no such specification.

The third limit concerns the boundary carried from the criterion’s provenance. Constraint 4.5 established that the transfer from the instructional setting leaves behind the presupposition of a trajectory toward self-regulation, so the test classifies an arrangement as it presently stands. The consequence is that the criterion issues no prediction:

an arrangement returning an extractive assignment today may return a generative assignment after a period of operation, and the account contains nothing establishing the conditions under which that transition occurs. A theory of the transition would be the natural extension of this paper, and it is absent from it.

A fourth consideration bears on interpretation in place of application. Claim 8.2 established that the capacity by which a relation holds itself in operation is produced relationally, through the institutions, markets, networks, and administrations available in a setting. An extractive assignment therefore reports jointly on the arrangement and on the relational infrastructure surrounding it, and the account supplies no operation separating the two contributions. Every reading of an extractive result as a finding about the arrangement alone accordingly overstates what the criterion returns.

10.2 Transfers Among Levels of Organisation Left Outstanding

The evidence assembled in this paper was obtained at three levels of organisation, and the paper holds each result at the level of its obtainment. This subsection states the three transfers the account leaves outstanding, each of which would extend the paper’s reach and none of which the paper performs.

The visibility result stands at the interpersonal level. Claim 5.3 was established on diary and experimental evidence obtained in dyadic settings, and its extension to organisational and international arrangements was marked as a conjecture at the point of statement. The observations establishing or defeating the extension would compare arrangements differing in the salience of the transfer to the supported organisation, holding routing and assigned position constant, and would ascertain whether the relation between salience and sustained capacity observed among individuals appears among organisations.

The drift result stands at the level of the funding structure. Claim 6.3 concerns the rates at which a sector produces arrangements of each kind under a given enforcement structure, and its application to an individual organisation requires the further argument that a structural tendency governs a particular case. That argument is not made here, and its absence bars the use of Claim 6.3 in the assessment of any named organisation.

The participation result stands within supported employment. Claim 9.2 was established on randomised evidence in a single domain, and its extension to other domains of assistance was marked as a conjecture. The domains in which the extension is most plausible are those sharing the structural feature that produced the result, namely the availability of the target activity as a setting within which support may be supplied.

10.3 The Quantities the Account Leaves Unestimated

An account supplying a criterion invites questions about distribution, and the present account answers none of them. This subsection states four such questions and the reason each stands outside the paper’s scope.

The account offers no frequency. The proportion of support arrangements presently in operation that return an extractive assignment stands unestimated, and the estimate would require the application of the criterion across a sample constructed for the purpose. The absence bars any inference from this paper to the scale of the phenomenon it describes.

The account offers no ordering among the determinants. Proposition 5.6 states that routing, visibility, and assigned position jointly settle the result, and it establishes no relative weight among them. A design facing a trade-off among the three accordingly obtains no guidance from the proposition, and the establishment of an ordering would require experimental variation of each determinant against the others.

The account offers no estimate of cross-scale return. Constraint 7.4 fixed the crossscale axis as an observation carrying no frequency and no account of selection, and the limit is repeated here because the axis is the point at which the paper’s language is most liable to be read as promising more than it establishes.

The account offers no population-level assessment. The Matthew-effect result reported in Section 3.4 established that capacitating provision distributes according to the capacity to take it up, so an arrangement passing the fading test for each participating relation may increase inequality across the population from which participants are drawn. The criterion operates at the level of the arrangement, and a population-level assessment requires a separate instrument the paper does not supply.

10.4 The Falsification Conditions

The principal claims of the paper attach to observations that would defeat them, and this subsection assembles those conditions in one place.

Claim 10.1. (Assembled falsification conditions). Proposition 5.6 is defeated by a set of arrangements matched on routing, visibility, and assigned position while separating systematically at withdrawal, which would establish an omitted determinant. Claim 5.2 is defeated by a set of arrangements separating on routing while returning matched results at withdrawal. Claim 5.5 is defeated by a demonstration that arrangements identical in the three determinants separate at withdrawal according to the supplying party’s stated purposes. Claim 6.3 is defeated by matched distributions of arrangement kind across organisations funded under structures differing in the strength of downward accountability, holding sector and setting constant. Claim 9.2 is defeated by randomised evidence in which preparation conducted in advance of placement produces higher rates of sustained participation.

Two elements of the account stand outside Claim 10.1 and their standing is stated here. Criterion 4.3 is a definition, so observation bears on its usefulness in place of its truth: the criterion is given up where its application proves impracticable across the settings it was introduced to serve, or where the classification it returns proves to carry no consequences for the design questions it was introduced to inform. Constraints 8.5 and 8.7 are normative, so observation bears on them at no point, and their standing rests on the arguments of Section 8 alone.

10.5 Summary of Results

This section established the following. Application of the criterion by actual withdrawal is barred wherever the supported party’s standing depends on the support, so application proceeds through matched comparison, planned transition, or inspection of the three determinants, with the third route carrying the practical weight and inheriting the risk attaching to Proposition 5.6. The account specifies the proportion of a transfer whose removal a relation absorbs and leaves unspecified the degree of function a relation must retain to count as holding. The criterion classifies present arrangements and issues no prediction of transition, and a theory of transition stands as the natural extension of the paper. An extractive assignment reports jointly on the arrangement and on the relational infrastructure surrounding it, with no operation separating the two contributions. Three transfers among levels stand outstanding: the visibility result beyond dyads, the drift result to individual organisations, and the participation result beyond supported employment. Four quantities stand unestimated: the frequency of extractive arrangements, the ordering among the determinants, the frequency and selection of cross-scale return, and the population-level distribution of capacitating provision. The falsification conditions attaching to the five principal empirical claims were assembled (Claim 10.1), and the definitional and normative elements of the account were identified as standing outside them. Section 11 summarises the results of the paper.

11 Summary of Results

The role of this section is to state the results of the paper as a whole, to fix the contribution at the width the preceding sections support, and to identify the work the account leaves to be done. The section states the results in the order of their establishment, then the contribution, then the outstanding work. Its method is recapitulation, and every result stated here carries the section at which its warrant was supplied.

The dependency claim was examined and reformulated. The proposition that transfers as a class produce durable reductions in recipients’ own effort was found without support in the randomised and review evidence, and the variation the same evidence reports across designs located the phenomenon in the structure of arrangements. Dependency was accordingly fixed throughout the paper as a property of the design and incentive structure of a support arrangement, ascertained arrangement by arrangement, and three provisions were adopted against the label’s documented allocative use (Section 2).

The territory already occupied was mapped. Five independent literatures were found to hold the distinction between support maintaining a recipient’s position and support building a recipient’s capacity: the intergroup helping literature with experimental support, Japanese development cooperation doctrine with its associated evaluation practice, human security doctrine, the social investment paradigm with its tested Matthew-effect critique, and the ownership-design literature with case material. Each was found to supply an operation of assignment serving its own purposes, and each such operation was found to require either a completed intervention or an inspectable design document.

The relational unit of evaluation was conceded alongside the distinction (Section 3).

The criterion was supplied. Support was defined as a transfer whose recipient is a relation, and the two kinds were defined by the dependence of the supported relation’s operation on continuation of the transfer. The fading test was stated as the operation ascertaining kind: a support is generative where its withdrawal leaves the supported relation intact, and extractive where its withdrawal collapses the relation it sustained. Fadeability was established as a graded property assessable by staged partial withdrawal. The test’s provenance was traced to the fading component of instructional scaffolding, and the transfer was bounded by the constraint that the instructional presupposition of a trajectory toward self-regulation remains behind. Outcome magnitude, duration of commitment, and expressed satisfaction were each shown to vary independently of kind, and the folding criterion was stated as the destination-side companion of the fading test (Section 4).

The determinants of kind were established. Routing, defined as the identity of the relation a transfer strengthens, was established analytically as a determinant. Visibility was established as a determinant at the interpersonal level on diary and experimental evidence. Assigned position, ascertained by the allocation of problem definition, method selection, execution decisions, and accounting, was established through the intergroup helping and helper-therapy literatures. The volume of a transfer, the medium through which it flows, and the purposes of the supplying party were established as inert with respect to kind (Section 5).

The prevalence of extractive arrangements was explained structurally. Accountability in the assistance sector was found asymmetric in enforcement, and the asymmetry was traced to all three determinants along separate paths: the attribution requirement favours routing to the supplier, the verification and identification requirements raise the salience of the transfer, and the results framework assigns the supported party a receiving position. The observation the fading test requires was found to fall beyond the horizon at which reporting terminates, and the value generated within programme relations was found to converge on the parties standing in the upward relation (Section 6).

The forms of return were separated and the paper’s claim among them was fixed.

Generative return was defined by its effect on the capacity of relations to generate further, and three independent axes were separated: time, subject, and scale. The first two were conceded to the theory of indirect and upstream reciprocity. The cross-scale axis was entered as the paper’s addition, distinguished by the arrival of a return of a different type at a level of organisation above that of the transfer, and fixed as an observation carrying no formal treatment, no frequency, and no account of selection, and competing with the diffusion and institutional literatures (Section 7).

The constraint on the criterion’s use was established. Worth was grounded through the separation of the source of value from the manner of valuing, so that value with a relational source stands as final value, fixed in advance of any criterion. Receipt was established as carrying no obligation of return. The argument grounding protection in the generative capacity of the system as a whole was stated in its strongest form, identified as utilitarian in structure, and declined as a ground while retained as an observation. The criterion was fixed as diagnostic, which answers the objection that a self-maintenance test cannot yield an evaluative conclusion. The lexical firewall was stated in exceptionless form, the limit case of support admitting no withdrawal was treated with both available errors named, and the deservingness firewall was stated with its mechanism half and its ethical half distinguished by register (Section 8).

The implications for design were drawn as predictions. An arrangement admitting fading routes its transfer to the counterparties of the supported relation, conducts staged partial withdrawals during operation, and establishes the resources, capacities, and linkages the relation will require in advance of withdrawal. Placement within an activity with support supplied inside it was found to produce higher sustained participation than preparation conducted in advance. The Japanese self-reliance system supplied the documented failure mode, in which the collapse of three senses of self-reliance into exit from benefit receipt aligns a system’s measured success with the reduction of its own expenditure, and the caution was carried to the present criterion. The accountability structure was identified as the lever (Section 9).

The limits were stated. Application by actual withdrawal is barred wherever the supported party’s standing depends on the support, so application proceeds by matched comparison, planned transition, or inspection of the determinants. The degree of function a relation must retain to count as holding stands unspecified. The criterion issues no prediction of transition. An extractive assignment reports jointly on the arrangement and on the relational infrastructure surrounding it. Three transfers among levels of organisation stand outstanding, four quantities stand unestimated, and the falsification conditions attaching to the five principal empirical claims were assembled (Section 10).

The contribution of the paper is accordingly narrow and is stated at that width.

The relational turn is held elsewhere. The extractive and generative distinction is held in five places. The dependency claim is contested and its general form stands unsupported. What the paper adds is an operation applying to a support arrangement while that arrangement is in operation, taking as its object the relation the arrangement maintains and returning a graded assignment; an account of the three features settling that assignment together with the three features standing inert with respect to it; a structural explanation of the prevalence of arrangements of the maintaining kind, operating through the same three features; one axis of return on which existing reciprocity theory is silent, entered as an observation; and a firewall holding the whole apparatus clear of allocation.

Three lines of work follow. The first is a theory of transition, specifying the conditions under which an arrangement returning an extractive assignment comes to return a generative one, which the criterion’s present form leaves entirely open. The second is the experimental separation of the three determinants, which would supply the ordering among them that design under trade-off requires. The third is the formal treatment of the cross-scale axis, which would specify the conditions under which a transfer at one level of organisation produces a return of another type at another, and would establish what such an account adds to the existing treatments of diffusion and institutional convergence.

References

Roger T. Ames. Confucian Role Ethics: A Vocabulary. University of Hawai’i Press, Honolulu, 2011.

Elizabeth S. Anderson. What is the point of equality? Ethics, 109(2):287–337, 1999.

Abhijit V. Banerjee, Rema Hanna, Gabriel E. Kreindler, and Benjamin A. Olken. Debunking the stereotype of the lazy welfare recipient: Evidence from cash transfer programs. The World Bank Research Observer, 32(2):155–184, 2017.

Francesca Bastagli, Jessica Hagen-Zanker, Luke Harman, Valentina Barca, Georgina Sturge, Tanja Schmidt, and Luca Pellerano. Cash transfers: What does the evidence say? Technical report, Overseas Development Institute, 2016.

Niall Bolger and David Amarel. Effects of social support visibility on adjustment to stress: Experimental evidence. Journal of Personality and Social Psychology, 92(3): 458–475, 2007.

Niall Bolger, Adam Zuckerman, and Ronald C. Kessler. Invisible support and adjustment to stress. Journal of Personality and Social Psychology, 79(6):953–961, 2000.

Giuliano Bonoli, Bea Cantillon, and Wim Van Lancker. Social investment and the matthew effect: Limits to a strategy. The Uses of Social Investment, 2017.

Robert Boyd and Peter J. Richerson. The evolution of indirect reciprocity. Social Networks, 11(3):213–236, 1989.

John Braithwaite. Crime, Shame and Reintegration. Cambridge University Press, Cambridge, 1989.

Tom Burns, Jocelyn Catty, Thomas Becker, Robert E. Drake, Angelo Fioritti, Martin Knapp, Christoph Lauber, Wulf Rössler, Toma Tomov, Jooske van Busschbach, Sarah White, and Durk Wiersma. The effectiveness of supported employment for people with severe mental illness: A randomised controlled trial. The Lancet, 370(9593): 1146–1152, 2007.

J. Baird Callicott. In Defense of the Land Ethic: Essays in Environmental Philosophy. State University of New York Press, Albany, 1989.

Bea Cantillon and Wim Van Lancker. Social investment and the matthew effect: Limits to a strategy. In Anton Hemerijck, editor, The Uses of Social Investment. Oxford University Press, Oxford, 2017.

Maia Chankseliani. International development and higher education: Two sides of a coin. International Journal of Educational Development, 94:102650, 2022.

Commission on Human Security. Human security now. Technical report, United Nations, New York, 2003.

Edward L. Deci, Richard Koestner, and Richard M. Ryan. A meta-analytic review of experiments examining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin, 125(6):627–668, 1999.

Paul J. DiMaggio and Walter W. Powell. The iron cage revisited: Institutional isomorphism and collective rationality in organizational fields. American Sociological Review, 48(2):147–160, 1983.

Alnoor Ebrahim. Accountability in practice: Mechanisms for ngos. World Development, 31(5):813–829, 2003.

Alnoor Ebrahim. Accountability myopia: Losing sight of organizational learning. Nonprofit and Voluntary Sector Quarterly, 34(1):56–87, 2005.

Michael Edwards and David Hulme, editors. Beyond the Magic Bullet: NGO Performance and Accountability in the Post-Cold War World. Kumarian Press, West Hartford, 1996.

Gøsta Esping-Andersen, Duncan Gallie, Anton Hemerijck, and John Myles. Why We Need a New Welfare State. Oxford University Press, Oxford, 2002.

David K. Evans and Anna Popova. Cash transfers and temptation goods. Economic Development and Cultural Change, 65(2):189–221, 2017.

Martha Albertson Fineman. The vulnerable subject: Anchoring equality in the human condition. Yale Journal of Law and Feminism, 20(1):1–23, 2008.

Jeffrey D. Fisher, Arie Nadler, and Sheryle Whitcher-Alagna. Recipient reactions to aid. Psychological Bulletin, 91(1):27–54, 1982.

Fourth High Level Forum on Aid Effectiveness. Busan partnership for effective development co-operation, 2011.

Anthony Giddens. The Third Way: The Renewal of Social Democracy. Polity Press, Cambridge, 1998.

Jacques T. Godbout and Alain Caillé. L’esprit du don. La Découverte, Paris, 1992.

Government of Japan. Official development assistance charter, 1992.

Cabinet decision; revised 2003.

Paul Harvey and Jeremy Lind. Dependency and humanitarian relief: A critical analysis. HPG Report 19, Overseas Development Institute, 2005.

Virginia Held. The Ethics of Care: Personal, Political, and Global. Oxford University Press, Oxford, 2006.

Anton Hemerijck. Changing Welfare States. Oxford University Press, Oxford, 2013.

Hideto Ikeya. 生活保護自立支援プログラム導入時の議論と到達点: 三つの自立が生活保護行政に与えた影響. 大原社会問題研究所雑誌, (717):29– 45, 2018.

Hans Jonas. The Phenomenon of Life: Toward a Philosophical Biology. Harper and Row, New York, 1966.

Michael B. Katz. The Undeserving Poor: From the War on Poverty to the War on Welfare. Pantheon, New York, 1989.

Marjorie Kelly. Owning Our Future: The Emerging Ownership Revolution. BerrettKoehler, San Francisco, 2012.

Eva Feder Kittay. Love’s Labor: Essays on Women, Equality, and Dependency. Routledge, New York, 1999.

Christine M. Korsgaard. Two distinctions in goodness. The Philosophical Review, 92 (2):169–195, 1983.

Alasdair MacIntyre. Dependent Rational Animals: Why Human Beings Need the Virtues. Open Court, Chicago, 1999.

Avishai Margalit. The Decent Society. Harvard University Press, Cambridge, MA, 1996.

Marcel Mauss. Essai sur le don: Forme et raison de l’échange dans les sociétés archaïques. Année Sociologique, 1925.

Thaddeus Metz. A Relational Moral Theory: African Ethics in and beyond the Continent. Oxford University Press, Oxford, 2022.

Ministry of Foreign Affairs of Japan. Development cooperation charter, 2015.

Revised 2023.

Ministry of Health, Labour and Welfare of Japan. 平成17年度における自立支援プログラムの基本方針について, 2005. 社援発第0331003号, Director-General Notice, Social Welfare and War Victims’ Relief Bureau. Ministry of Health, Labour and Welfare of Japan. 生活困窮者自立支援制度, 2015.

Act of 2013, in force 2015.

G. E. Moore. Principia Ethica. Cambridge University Press, Cambridge, 1903.

Nathalie Morel, Bruno Palier, and Joakim Palme, editors. Towards a Social Investment Welfare State? Ideas, Policies and Challenges. Policy Press, Bristol, 2012.

Arie Nadler. Inter-group helping relations as power relations: Maintaining or challenging social dominance between groups through helping. Journal of Social Issues, 58(3): 487–502, 2002.

Arie Nadler and Samer Halabi. Intergroup helping as status relations: Effects of status stability, identification, and type of help on receptivity to high-status group’s help. Journal of Personality and Social Psychology, 91(1):97–110, 2006.

Nel Noddings. Caring: A Feminine Approach to Ethics and Moral Education. University of California Press, Berkeley, 1984.

Martin A. Nowak and Karl Sigmund. Evolution of indirect reciprocity by image scoring. Nature, 393:573–577, 1998.

Martin A. Nowak and Karl Sigmund. Evolution of indirect reciprocity. Nature, 437: 1291–1298, 2005.

Martha C. Nussbaum. Frontiers of Justice: Disability, Nationality, Species Membership. Harvard University Press, Cambridge, MA, 2006.

Organisation for Economic Co-operation and Development. The paris declaration on aid effectiveness, 2005.

With the Accra Agenda for Action, 2008.

Roland Paris. Human security: Paradigm shift or hot air? International Security, 26 (2):87–102, 2001.

Tom Regan. The Case for Animal Rights. University of California Press, Berkeley, 1983.

Frank Riessman. The “helper” therapy principle. Social Work, 10(2):27–32, 1965.

Beatrice Lorge Rogers and Jennifer Coates. Sustaining development: A synthesis of results from a four-country study of sustainability and exit strategies. Technical report, FANTA/FHI 360, for USAID, 2015.

Everett M. Rogers. Diffusion of Innovations. Free Press, New York, 5 edition, 2003.

Social Security Council, Japan. 生活保護制度の在り方に関する専門委員会報告書, 2004.

Richard M. Titmuss. The Gift Relationship: From Human Blood to Social Policy. Allen and Unwin, London, 1970.

Wim van Oorschot. Who should get what, and why? on deservingness criteria and the conditionality of solidarity among the public. Policy and Politics, 28(1):33–48, 2000.

Lev S. Vygotsky. Mind in Society: The Development of Higher Psychological Processes. Harvard University Press, Cambridge, MA, 1978.

Tetsurō Watsuji. Rinrigaku. Iwanami Shoten, Tokyo, 1937.

David Wood, Jerome S. Bruner, and Gail Ross. The role of tutoring in problem solving. Journal of Child Psychology and Psychiatry, 17(2):89–100, 1976.