Paper

Abstract

Paper 2

An earlier version of this study concluded that five seminal theories in the field of philanthropy share three “unspoken assumptions”: (renewal of the act of giving, the actor remaining alive, and the beneficiary existing at the time of the decision). This paper re-examines that claim through the original texts of Andreoni, Weisbrod, Hansmann, Salamon, and Salamon & Anheier, after disaggregating the third assumption into two distinct questions: (1) the representation of future benefit or a later beneficiary, and (2) the creation of an enduring legal right with a rights-holder and a rule for enforcement under Islamic law and through the courts. It uses four-way coding that distinguishes between: (explicit, textually entailed, outside the model’s scope, and untested). The earlier claim does not withstand examination:

  • Representation of the future: explicit in three theories, untested in a fourth, and outside the level of analysis of the fifth.

  • Creation of an enduring right: none of the theories offers a positive representation of the creation of an enduring legal right. Yet this absence has no common cause: it lies outside the scope of three theories and remains untested in two.

Thus neither a shared structural incapacity nor the uniqueness of waqf (a legal-institutional arrangement in which property is dedicated to a designated purpose under a founder’s stipulations and managed by an administrator) is established. The contribution here is therefore not to establish a new gap, but to explain that “silence takes different forms”, and that conflating levels of analysis turns the limits of a question into a defect attributed to a theory. Waqf becomes a probe for refining the question, rather than a case that defeats the theories.

Keywords: theories of philanthropy; model boundaries; waqf; future beneficiary; enduring rights; textual coding.

1. The Problem: When the Conclusion Preceded the Question

Paper (P124) began with a valid observation: waqf has a limited presence in the seminal studies of philanthropy. It then made an inferential leap from the absence of discussion to a stronger claim: that the five theories cannot represent waqf because they share three “unspoken assumptions”.

These assumptions were formulated around three intrinsic characteristics of waqf: (a founding act that does not recur, a founder’s restriction that continues after the founder’s death, and rights for generations not yet born). Waqf thus became a test, and the failure to find its characteristics in each theory became evidence of that theory’s incapacity.

Yet the analytical framework combined three errors in a single question:

First: disregarding differences in levels of analysis

The five theories do not operate at the same analytical level: (1) Andreoni explains the individual’s decision to give, while Weisbrod explains how the production of collective goods is distributed across sectors; (2) Hansmann explains the existence of the non-profit form in relationships lacking adequate contractual safeguards; (3) Salamon explains voluntary failure and the relationship of voluntary provision with government; and (4) Salamon and Anheier explain the sector’s size and structure through national historical patterns. The question of creating a legal right within an individual unit is therefore not absent in the same way from a model of the individual and a theory of an entire social system.

Second: conflation within a unified concept of “the future”

“The future” was not a single variable. A benefit directed towards a coming generation is one thing; an institution selecting a beneficiary after the funder’s decision is another; and the creation of a legal right for a future group, with a person with standing and an enforcement rule, is a third, distinct matter. The earlier coding assigned the same blanket judgement to all three dimensions.

Third: reducing different forms of textual silence and omission to one

Textual silence and omission do not warrant a single blanket judgement. An element may lie outside a theory’s stated purpose and level of analysis; it may fall within its scope but remain empirically untested; it may be entailed by the text’s construction without being explicitly named; or it may be explicitly stated. When these cases are grouped under the bare label “absent”, agreement among the theories appears broader than it actually is.

This paper therefore does not ask: why are the theories incapable of recognising waqf? It asks a narrower, more readily testable question: “How does the assessment of five theories of philanthropy change when we separate the level of analysis, the representation of the future, and the creation of a legal right, and distinguish being outside scope from being untested?”.

The central thesis: theoretical silence is not a single phenomenon. The non-representation of an element cannot properly be treated as structural incapacity before establishing where that element falls in relation to the theory’s scope.

2. What Remains of the Earlier Work and What Is Withdrawn

Epistemological rigour requires an explicit disaggregation of what is retained and what is withdrawn from the earlier preliminary reading (P124):

a. What remains of the earlier work (four points): first, the selection of the five texts as a purposive sample of different explanatory approaches, rather than a comprehensive survey of the field; second, the use of waqf as a probe that tests the limits of concepts; third, the distinction between criticism of a theory and judgement of an institution; and fourth, the recognition that the absence of the name waqf from a text does not, in itself, indicate its rejection or the impossibility of representing it.

b. Four items are withdrawn (four claims): (1) the claim that the theories share three assumptions; (2) the attribution of structural incapacity to all five theories collectively; (3) the claim that waqf “breaks” them because it combines characteristics found together nowhere else, since the replacement test in P128 established that European and non-Islamic comparators share elements that the first version had considered unique to waqf; and (4) philosophical genealogies such as “Bentham → Andreoni”, unless supported by independent historical texts, since structural similarity does not establish intellectual lineage.

This withdrawal is not an editorial loss, but a gain and an advance. The narrower claim emerging from the examination is stronger than the earlier broad claim: a question that has not been disaggregated can manufacture a theoretical deficiency absent from the text by asking a model operating at one level to account for a construction situated at an entirely different level.

3. The Sample and the Matrix

3.1 The Five Theoretical Texts Examined

The following original texts were examined:

1. James Andreoni, the 1989 and 1990 papers on giving with a personal “warm glow”.

2. Burton A. Weisbrod, Toward a Theory of the Voluntary Non-Profit Sector in a Three-Sector Economy, in the 1972 IRP version preceding the published chapter.

3. Henry B. Hansmann, “The Role of Nonprofit Enterprise”, 1980.

4. Lester M. Salamon, “Of Market Failure, Voluntary Failure, and Third-Party Government”, 1987.

5. Lester M. Salamon and Helmut K. Anheier, Social Origins of Civil Society, the 1996 working version and the 1998 published version.

Four texts are held locally with their hashes, alongside Andreoni’s two papers. The Salamon and Anheier text was examined in an online copy posted by its authors, using the pagination of the 1996 working version, without a local file or a retained verbatim quotation; efforts to obtain the file have been exhausted. Details of the evidentiary record appear in THEORY-MATRIX/1.0 and the accompanying acquisition log.

3.2 The Four Dimensions

CodeQuestion
D1Does the model represent a founding decision to provide funding that requires no renewed giving by the founder?
D2Does it represent the continuation of the founder’s restriction after the founder’s death?
D3-WDoes it represent future benefit or a beneficiary identified after the funding decision?
D3-RDoes it represent the creation of an enduring legal right for a future group, with a rights-holder and an enforcement rule?

Splitting D3 is the governing methodological decision. An institution’s later selection of a beneficiary does not entail that the beneficiary is unborn; mentioning future generations in a utility function does not create a right for them; and the enforceability of a donor’s restriction does not identify who is entitled to demand its enforcement a century later.

3.3 The Four Codes

CodeDefinition
EThe element is explicitly stated in the text or a model assumption
IThe element is textually entailed, even if unnamed
OThe element lies outside the theory’s level of analysis or stated scope
UThe element belongs to the same analytical level at which the text operates, but the text does not test it sufficiently to permit a judgement

The codes are not degrees of confidence. I does not lie midway between E and U. Nor does O mean that the theory denies the element; it means that the theory does not claim to explain it.

3.4 The Rule Distinguishing Being Outside Scope (O) from Being Untested (U)

The second reading showed that the expression “in the vicinity of the theory” was too broad to define the boundary between the out-of-scope code (O) and the untested code (U). A more precise rule was therefore adopted: verbal or thematic proximity alone is insufficient; the question must belong to the same class of variables that the text addresses. The continued force of the donor’s restriction was thus untested in Hansmann, because he introduced enforceability into the text. By contrast, the creation of a right, the person with standing, and the enforcement rule fall outside Andreoni’s utility model and Weisbrod’s model of outputs and funding, even though they mention bequests or a social compact.

Nor is (I) assigned merely because an event is plausible or familiar. The construction of the text must make its denial inconsistent with that text. Salamon’s “third-party government” therefore does not entail that the beneficiary is identified after the funding decision. It establishes a distribution of decision-making authority between two bodies, not a necessary temporal sequence.

3.5 Independent Reading and Methodological Sensitivity Testing

a. Direct agreement rate and the adjudication process:

The sixteen cells relating to texts held locally were read independently on two occasions, in two stages separated by a period of examination. Raw agreement was (three of six) in the first stage and (five of ten) in the second: (eight of sixteen), or fifty per cent (50%), before adjudication.

This statistical rate is not presented as evidence of high reliability. Its immediate value lies in exposing the ambiguity at the boundary between (O) and (U). The preceding rule was therefore added, and the eight disagreements were resolved on that basis.

b. Sensitivity test results:

The result’s sensitivity to the last five disagreements was tested: (1) Matrix robustness: no column contained the same code across all five theories under any combination of the two readings. (2) Exclusion test for the weaker evidentiary record: the theoretical result holds when the Salamon & Anheier row, with its weaker evidentiary record, is excluded. The differences and diversity are already established in the four theories whose original texts are held locally.

3.6 The Study’s Methodological Falsifiers

The paper’s claim fails if any of three conditions emerges: the distinction between (O) and (U) cannot be defined by reference to the boundaries of the texts; the earlier judgement remains unchanged after D3 is split; or the theory operating at the macro level turns out to contain internal variables for creating institutional rights.

4. Findings

4.1 Andreoni: The Future Is Present, While the Institution Lies Outside the Question

Andreoni builds his model around an individual who allocates wealth between private consumption and giving towards a public good: xi + gi = wi. Both the quantity of the public good and the individual’s own gift enter the individual’s utility. An enduring founding endowment therefore lies outside the structure of the decision (D1=O), just as the restriction imposed by a deceased founder lies outside the unit of analysis (D2=O).

Yet the text is not silent about the future. The 1989 paper states: “Suppose, however, that parents feel an obligation to future generations” (p. 1456). This explicit assumption is part of the model and cannot be downgraded to an entailment or a possibility (D3-W=E). The enduring legal right, the claimant, and the enforcement rule, however, are not variables in the utility model (D3-R=O).

The finding is therefore not that Andreoni “does not see the future”, but that he represents it as a motive and a benefit without constructing a theory of institutional rights.

4.2 Weisbrod: An Intergenerational Contract Within a Static Model

Weisbrod states on the first page that his analysis is “essentially static” and that it explains the existence of organisations with a minimum of institutional assumptions. He examines the funding of voluntary production through current contributions, but does not test whether a permanent founding endowment could remove the need to renew those contributions. Because funding is central to the theory rather than outside it, the judgement is D1=U, not O.

The deceased founder’s restriction and the mechanism sustaining it lie outside this static construction with its minimum of institutional assumptions (D2=O). Nevertheless, the text mentions “future generations of young people” who finance the preceding generation’s retirement, and calls the relationship “a social compact” (p. 23). The future is therefore explicit (D3-W=E). Yet the text constructs neither a founding instrument nor a right or an enforcement rule (D3-R=O).

The effect of splitting the question is clear here: the theory represents time across generations, but does not represent the legal form that might hold a right across that time.

4.3 Hansmann: Restrictions and Later Beneficiaries Alongside Contract Failure

Hansmann examines the non-profit form as a response to the inability to monitor a producer or contract fully with it. He thus comes closer to the institutional dimensions than Andreoni or Weisbrod. He mentions a single donation made long ago that would have given its donor a permanent vote had the organisation taken a stock-based form, and notes that donors impose restrictions on purposes that are “often enforceable” (pp. 852–853). Yet he does not turn this into a model of a permanent founding endowment or test the continuation of the restriction after death. The judgements are therefore D1=U and D2=U.

He also explicitly describes a wealthy funder handing part of their wealth to a relatively independent organisation “that will itself choose the ultimate recipients” (p. 848 n.41). The ultimate beneficiary is thus identified after the funder’s decision (D3-W=E). But the combination of later selection and the enforceability of some restrictions neither entails a right for a future group nor identifies the person with standing (D3-R=U).

4.4 Salamon: The Right Is Present as a Limit to Charity

Salamon discusses the dependence of voluntary provision on contributions, the authority of resource-holders to determine what the sector does and whom it serves, and the “particularism” of the groups to which organisations respond. The first three questions therefore do not lie outside his scope, but they are not tested in their waqf form: no founding endowment that removes the need for renewed resources, no deceased founder’s restriction, and no future beneficiary (D1=U, D2=U, D3-W=U).

More significantly, he places the dimension of rights at the centre of the distinction between private charity and government: “Aid is provided as a matter of charity, not of right”, and he considers that the charitable system does not readily create a claim to assistance as a right (pp. 41–42). Rights therefore do not lie outside the theory, but an enduring right for a future group and the person entitled to enforce it remain untested (D3-R=U).

This judgement is more precise than describing the theory as one that “does not see rights”. It sees them, but in the contrast between charity and the state, rather than within an architecture of enduring private allocation.

4.5 Salamon and Anheier: A Different Level of Analysis

Social origins theory declares its unit of analysis to be: (the non-profit sector and its components), and selects the scale of employment and sources of support as its two principal variables. It explains the sector’s size and structure through its embeddedness in power relations, class relations, and patterns of national regimes. All four dimensions therefore lie outside its level of analysis: the decision to establish a founding endowment, the deceased founder’s restriction, the later beneficiary, and the creation of a right (O/O/O/O).

This assessment is not based on a judgement that this is the weakest theory at representing waqf; the theory answers a question at a different level. The effects of awqāf may enter the sector’s size and funding sources without the act of creating a waqf becoming a variable in the theory. Requiring the theory to address that act conflates explaining a sector’s structure with explaining the founding instrument of a unit within it.

⚠️ An evidentiary limitation governs this assessment: the four judgements for this theory rest on online examination without a local file or a verbatim quotation, and its page references follow the 1996 version. They are suggested by the stated boundaries of the unit of analysis, rather than settled by retained textual evidence, and the conclusion does not depend on them alone.

4.6 The Overall Result

TheoryD1D2D3-WD3-R
AndreoniOOEO
WeisbrodUOEO
HansmannUUEU
SalamonUUUU
Salamon and AnheierOOOO

The overall result: three E, nine O, eight U, and no I. No column carries the same code across all five theories. There is therefore no “shared unspoken assumption” as defined by the earlier version. This remains true even if the Salamon and Anheier row, with its weaker evidentiary record, is excluded: the differences within the columns across the four theories whose texts are retained are sufficient.

5. What Did Waqf Actually Reveal?

Using waqf as such did not reveal a single incapacity across five theories; it explicitly revealed four methodological boundaries that the earlier question had obscured and conflated:

1. It revealed the boundary of the unit of analysis: a question suited to a theory of the organisation is not suited, in the same form, to a theory of individual motivation or a theory of national regimes.

2. It revealed the boundary of time: mentioning the future does not create a permanent institution, and an intergenerational social compact is not a right in rem or a right in personam for a group that does not yet exist.

3. It then revealed the boundary crossed in moving from benefit to right: a utility function can represent concern for coming generations, but this does not identify who can make a claim, before whom, or for what remedy. This is less a gap within the function than a move from economics to legal architecture.

4. It revealed the boundary of waqf exceptionalism: P128 established that an enduring founding act, a founder’s restriction, and a later beneficiary are not unique to the institution of waqf. The appropriate probe is therefore not: “Does the theory recognise waqf?”, but: “Which elements of the family of purpose institutions fall within the question the theory was designed to answer?”.

Waqf’s role thus changes. It is no longer a “theory-breaking case”, but a probe for disaggregation: an institution with a complex combination of characteristics that compels the researcher to separate what they have habitually grouped together: (motivation, funding, entity, restriction, beneficiary, right, and enforcement).

6. Competing Interpretations

6.1 “The Result Is Merely a Linguistic Qualification Distinguishing O from U”

It may be argued that the distinction between O and U is terminological and changes nothing: the enduring right is absent in both cases. Yet the distinction changes the kind of critique and the appropriate response. If the element lies outside scope, the theory is not criticised for failing to explain it; it is paired with a theory operating at another level. If the element is untested within its scope, testing can be extended without claiming that the theory must be rebuilt. The distinction between complementing a theory and correcting it therefore depends on this coding.

6.2 “The Absence of Positive Representation Is Sufficient to Establish Deficiency”

A claim of theoretical deficiency first requires an agreed “functional standard”. A microscope cannot be described as deficient or incapable merely because it does not measure national income.

What can be stated precisely as an established finding is that the five theories, taken together in the versions examined, do not offer a positive theory of the creation of an enduring legal right. This is a gap in the coverage of the field’s collective toolkit, rather than a structural incapacity shared by every individual tool.

6.3 “The Choice of Waqf Still Directs the Result”

This is partly true, which is why the paper does not claim that its dimensions are exhaustive. The four dimensions are derived from a problem exposed by long-lived purpose institutions. Another institution may expose different dimensions. Yet the falsifier here is internal: even when these waqf-oriented dimensions were used, the texts did not yield the agreement claimed by the first version.

6.4 “The Collapse of the Earlier Claim Removes the Paper from the Series”

The paper belongs in the series because it answers the series’ constitutive question, rather than because an earlier finding survived: “Is the deficiency specific to waqf, shared by the family of purpose institutions, or located in the instrument of mapping and measurement itself?”.

Its answer: part of the deficiency lay in the researcher’s own instrument, specifically in conflating levels of analysis and their different questions under a code of absence. This is a necessary finding for a series whose central subject is: “Who draws the map?”.

7. The Methodological Contribution

The paper yields a small methodological tool that can be transferred and applied to any comparative critique of theories and models of knowledge:

1. Identify the unit of analysis: determine the theory’s unit of analysis before testing it against an institutional case.

2. Disaggregate the characteristic: turn a composite characteristic into questions that do not entail one another.

3. Use four-way coding: distinguish explicit statement, entailment, being outside scope, and being untested.

4. Assess the significance of the name: do not treat the absence of an institution’s name as evidence of the absence of its structure.

5. Apply a functional standard: attribute incapacity to a theory only in a function it claims to perform.

6. Combine models: when an element is distributed across several levels, what is needed is a combination of theories, rather than the demolition of one.

This tool may be called a “model boundaries profile”. It corresponds to the “estimate provenance profile” in the first paper of this series: the latter traces a number back through the sequence of its construction; the former traces a theoretical judgement back to the level of the question that produced it.

8. Methodological Limitations

The study undertakes to state its scholarly limitations and adhere strictly to them:

First: the sample comprises five selected original formulations, rather than a comprehensive bibliographical survey of subsequent developments.

Second: raw agreement between the two readings reached fifty per cent before adjudication. This reveals the sensitivity of the boundary between O and U, so the result cannot be presented as stable, automatic coding. This was addressed through an explicit rule, recorded adjudication, and a sensitivity test of the result.

Third: the Salamon and Anheier version was examined through an online copy posted by its authors, and its file was not retained. Its cells therefore have a lower evidentiary status, and no essential part of a claim depends on them. Fourth, the Weisbrod text examined is the 1972 version preceding the chapter published in 1975 and 1977. The formulations may differ, so the judgements are tied to the version examined.

Fifth: the paper tests neither the validity of the empirical theories nor their predictive power; it tests where four questions fall in relation to the boundaries of their texts.

Sixth: the absence of a positive theory of enduring rights among these five does not establish its absence from legal scholarship, the theory of the institution, or the literature on trusts and awqāf.

9. Conclusion

The earlier version’s error was not in seeing a gap where something was indeed missing, but in turning different gaps into a single gap. The future is explicitly present in three texts; rights are present as a limit to charity in Salamon; and the donor’s restriction is present and enforceable in Hansmann. Nevertheless, the five theories do not offer a positive model of the creation of an enduring legal right for a future group. The reason for this absence, however, differs from text to text and does not amount to a shared unspoken assumption.

The more consequential finding concerns the method of criticism. It is not enough to bring a complex institution to a theory and count what we do not find there. We must first ask: does it fall within the theory’s function? Has the theory explicitly described an analogue? Has it left the element untested? Have we moved from benefit to right, or from individual to sector, without acknowledging the transition?

In this sense, waqf served the paper by defeating the hastily formulated question, rather than the theories. This contribution is closer to the series’ purpose: not every empty cell indicates a deficiency in what is being mapped; it may indicate a deficiency in the mapping instrument.

Selected References

Andreoni, James. “Giving with Impure Altruism: Applications to Charity and Ricardian Equivalence.” Journal of Political Economy 97, no. 6 (1989): 1447–1458.

Andreoni, James. “Impure Altruism and Donations to Public Goods: A Theory of Warm-Glow Giving.” Economic Journal 100, no. 401 (1990): 464–477.

Hansmann, Henry B. “The Role of Nonprofit Enterprise.” Yale Law Journal 89, no. 5 (1980): 835–901.

Salamon, Lester M. “Of Market Failure, Voluntary Failure, and Third-Party Government.” Journal of Voluntary Action Research 16 (1987): 29–49.

Salamon, Lester M., and Helmut K. Anheier. “Social Origins of Civil Society: Explaining the Nonprofit Sector Cross-Nationally.” Voluntas 9, no. 3 (1998): 213–248.

Weisbrod, Burton A. Toward a Theory of the Voluntary Non-Profit Sector in a Three-Sector Economy. Institute for Research on Poverty Discussion Paper, 1972; published in 1975 and revised in 1977.

The governing audit record: THEORY-MATRIX/1.0, hash d9ff45a96246cb7393633a4e0080618bef06c4db5ed4010a46458aa8012bb3aa. THEORY-MATRIX/0.10 preserves the record of the two readings before adjudication.