← Atlas
Idea

Governance, stewardship, and accountability

The Hershey School Trust's 2002 effort to sell its controlling Hershey Foods stake exposes a bounded governance conflict: trustees pursued diversification, Pennsylvania's attorney general invoked public supervision, and a court preserved the status quo without deciding the merits. Independent accounts disagree over whether intervention protected community accountability or imposed costs on the school's beneficiaries.

Governing questionWhen governors claim they are protecting an institution's purpose, who is entitled to demand an account—and who can stop them?

Working · Claim Cited

A perpetual instrument distributed authority

Milton and Catherine Hershey's 1909 deed created a perpetual school, named a trustee and managers, assigned investment decisions to the trustee with manager approval, required operational and financial records, and supplied a method for filling manager vacancies. The version available from the school is the Second Restated Deed approved in 1976, not a facsimile of the unamended 1909 instrument.1 It establishes formal offices and powers through that date; it does not establish how later officeholders actually deliberated or how students experienced their decisions.

The controlling company interest arrived later. The 2002 judicial record says that a controlling interest in Hershey Foods had comprised the school trust's principal corpus since 1918. It also records that the same directors served as school managers and that the trust held 77 percent of the company's voting power during the sale dispute.2 The restated deed does not itself document the 1918 stock transfer, so the deed and court record perform different evidentiary roles.

This arrangement joined a charitable purpose, a school-governance office, investment authority, and corporate control. It did not make purpose self-executing. Managers still had to interpret how investment risk, school support, company control, and succession fit together. The deed gave managers substantial authority; it did not give current students, alumni, employees, shareholders, or residents an electoral route into that office.1 Formal authority and affected-party standing therefore remained different questions.

The proposed sale produced competing stewardship claims

In 2002 the trust initiated a process to sell its controlling interest in Hershey Foods in order to diversify trust assets. The attached trial adjudication described the company stake as 58 percent of trust asset allocation and 77 percent of company voting power; it also recorded the trustees' diversification rationale and the impending bid process.2 These are facts and positions in a preliminary-relief record, not a final finding that a sale was prudent or imprudent.

Mark Sidel's independent legal history reconstructs rapid opposition by company management, employees, residents, alumni, legislators, and the attorney general. It traces how the dispute moved from portfolio policy to competing definitions of beneficiary and public interest, and how the trust terminated the sale process after the injunction.3 Sidel's account is valuable because it assembles legal filings and contemporaneous reporting close to the events. It is still a scholar's interpretation, not a participant-controlled archive or a representative sample of the people affected.

Jonathan Klick and Robert Sitkoff supply a sharply different independent reading. Their event study associates the sale announcement with a positive abnormal stock return above 25 percent and cancellation with a negative abnormal return near 12 percent. They interpret those movements as evidence of agency costs, lost diversification, and reduced value for the trust and shareholders.4 Their own design is a single-firm event study and cannot separate every alternative explanation. Stock-price response also does not measure student welfare, employment security, community continuity, or the distribution of gains and losses.

The disagreement cannot be collapsed into one settled lesson. Sidel emphasizes community accountability and the political representation of interests that might otherwise be excluded. Klick and Sitkoff emphasize portfolio risk, capital-market valuation, and the possibility that political intervention injured the named charitable purpose. Both evaluate the same interruption through different welfare measures and different accounts of who should count.

The injunction preserved review without deciding the merits

Pennsylvania's attorney general petitioned the Dauphin County Orphans' Court, which temporarily barred the trust from entering an agreement or understanding that could commit it to a sale. The Commonwealth Court affirmed because it could not find that the lower court lacked reasonable grounds for preliminary relief. It directed an expedited merits decision because a completed sale would have made review irreversible.2 That holding preserved the status quo; it did not finally decide whether the attorney general could direct the trust's investment process or whether a particular sale should occur.

The attached trial adjudication treated potential harm to employees and Derry Township as relevant to the public interest. Judge Pellegrini's dissent argued that intervention before a final agreement risked making the attorney general a co-manager and exceeded the statutory boundary of review.2 The primary record therefore contains both the public-supervision rationale and an institutional objection to the reviewing forum's reach.

A 2006 Pennsylvania Supreme Court decision adds a separate standing boundary. It concerned an alumni challenge to the modification of an administrative agreement, not the abandoned sale. The court held that the alumni association lacked the special interest required to enforce the trust and pointed instead to the attorney general's enforcement authority.5 Alumni advocacy and legal standing were not equivalent, but the later holding should not be read as a merits judgment about the 2002 sale.

These proceedings support a narrow proposition: accountability can depend on an office capable of pausing an irreversible action, while the authority and representativeness of that office remain contestable. They do not establish that judicial review always improves governance or that informal voice is irrelevant when it lacks legal standing.

Agency and stewardship are comparative models

Jensen and Meckling define an agency relationship as a contract in which one person engages another to perform a service and delegates decision authority. Their model treats monitoring expenditures, bonding expenditures, and residual loss as agency costs.6 Applied comparatively, it asks what discretion trustees held, how their conduct was monitored, and which divergence remained after monitoring.

The model does not choose the proper principal in a charitable trust. Current students, future eligible children, the public, shareholders, and people exposed to corporate decisions occupy different legal and material positions. Treating one as the principal can make the others disappear analytically. No source reviewed here establishes that Jensen and Meckling influenced the 2002 actors.

Davis, Schoorman, and Donaldson contrast agency assumptions with a stewardship model in which managers may identify with collective objectives and find pro-organizational conduct intrinsically rewarding. Their framework makes psychological and situational conditions, rather than universal opportunism, part of governance design.7 It does not show that a claimed steward is faithful, or that identification with an organization includes every person and ecosystem affected by it.

Agency and stewardship therefore raise rival diagnostic questions. Neither infers motive from office, and neither adjudicates the sale. A trustee can invoke mission while misjudging consequences; a monitor can invoke beneficiary protection while pursuing a narrower political or financial measure.

Accountability is one actor–forum framework

Bovens uses accountability in a deliberately narrow sense: an actor must explain and justify conduct to a forum that can question, judge, and expose the actor to consequences. He separately evaluates arrangements through democratic, constitutional, and learning perspectives.8 The framework helps distinguish disclosure from an answerable relationship.

Hershey contained several incomplete forums. The trust and school boards held interlocking roles. Public shareholders could not outvote the controlling interest. Employees, residents, and alumni could organize and provide information without thereby gaining trust-enforcement rights. The attorney general and courts could intervene, but the dissent and later standing case show that their boundaries were disputed.25

Bovens's definition does not establish which forum deserves authority, whether its questions use the right evidence, or whether consequences improve the underlying institution. An actor–forum relation can exclude affected parties, punish candor, or substitute one unreviewed governor for another. Explanation, questioning, judgment, and consequences are observable features, not a complete theory of legitimacy.

Affected-party and ecological evidence remains incomplete

The legal and scholarly record gives students a powerful beneficiary position but offers little direct student testimony about the 2002 choice. Community and employee positions appear chiefly through court findings, filings, and Sidel's use of contemporary reporting. The current alumni association describes its mission as supporting alumni community and the deed's principles, but that present self-description does not reconstruct its 2002 position or represent current students, employees, or Derry Township residents.9

The 2002 dispute also centered on a Pennsylvania trust and town while the company operated through wider labor and ingredient systems. Hershey's 2025 responsible business report identifies current sourcing, human-rights, workforce, and environmental topics in the company's own reporting. A U.S. Department of Labor-funded NORC study independently documents hazardous child labor across the cocoa sectors of Côte d'Ivoire and Ghana.10 Neither source attributes a particular 2002 Hershey governance decision to a measured supply-chain outcome: the company report is a later self-report, and the NORC study is sector-wide rather than company-specific.

No selected source is controlled by current students, production workers, cocoa farmers, supplier workers, or cocoa-growing communities. None gives nonhuman life, ecosystems, or future generations a direct forum in the 2002 decision. Those gaps prevent an impact classification and make it inappropriate to treat the court-recognized public interest as a complete account of everyone affected.

Structured research record

The seven nodes separate the case from later analytical models:

  • I14-N01 is the formal trust arrangement documented by the deed and court record; I14-N02 is the 2002 sale process and its diversification rationale. 12
  • I14-N03 is attorney-general and judicial review, including the preliminary injunction and the dissent over institutional reach.2
  • I14-N04 is Jensen and Meckling's agency-cost model; I14-N05 is Davis, Schoorman, and Donaldson's stewardship model.67
  • I14-N06 is Bovens's narrow actor–forum concept.8
  • I14-N07 is an editorial diagnostic joining authorization, evidence, standing, review, remedy, and succession. It is not a doctrine found in the deed, either judicial opinion, or the three theory articles.

The six relations have bounded status and support:

  • I14-E01 and I14-E02 reconstruct the formal investment authority, sale process, attorney-general petition, and preliminary restraint. Both attach the internal Hershey record and reject broader conclusions about final merits or universal community standing.12
  • I14-E03 applies agency theory comparatively; I14-E04 applies stewardship theory comparatively. They attach the Hershey case for internal navigation, while the external theory originals remain in their respective notes. Neither asserts influence or observed motive.67
  • I14-E05 applies Bovens comparatively to the review forums. It does not claim that his framework shaped the proceeding or selected the legitimate forum.8
  • I14-E06 organizes the deed, dispute, review, and later standing decision into an editorial chain. It is not a judicial holding, an actor's documented model, or a general causal sequence.25

The four Workloop records remain proposals. I14-P01 proposes binding each consequential coworker to a responsible principal; I14-P02 proposes narrow, inspectable, revocable grants; I14-P03 proposes durable records of decisions, approvals, dissent, conflicts, and exceptions; and I14-P04 proposes succession and incident-recovery rules. Each record specifies evidence needed and an abuse question. None is validated by Hershey, agency theory, stewardship theory, or Bovens.

There is no structured impact record and no reading-dependency record. Impact is unclassified rather than absent, and no prerequisite reading order is asserted. There are also no typed idea or work ID arrays; the three destinations below remain in the untyped related set. These absences keep an editorial comparison or product hypothesis from becoming a measured outcome, documented intellectual influence, or required sequence.

Linked analytical coordinates

The related set contains three internal routes: The Hershey Company and Milton Hershey School Trust, Newman's Own, and the DAO ideal. These are editorial navigation coordinates. They do not establish shared legal form, equivalent beneficiaries, common outcomes, or historical descent.

The 112 institution coordinates are broader prompts for comparing authorization, office, evidence, standing, review, remedy, and succession. Their presence does not classify any institution as legitimate, accountable, successful, failed, or equivalent to Hershey.

Historical, military, industrial, and corporate coordinates:

Comparative coordinates from ACOFOP through FUCVAM:

Comparative coordinates from Fujimori through Northern Rangelands Trust:

Comparative coordinates from Odebrecht through CAMPFIRE:

Each institution supplies its own evidentiary record. Its coordinate here does not establish the direction, legitimacy, effectiveness, harms, beneficiary definition, or affected-party standing of its governance.

Evidence still needed

  • The trustees' 2002 minutes, investment advice, bid materials, internal alternatives, conflicts disclosures, and individual votes, sufficient to test the rationales reported in litigation and scholarship.
  • Contemporaneous records controlled by students, alumni, employees, unions, minority shareholders, local businesses, and residents rather than accounts mediated principally through officials, courts, or journalists.
  • Company- and decision-specific evidence tracing any sale or non-sale effects through employment, school services, investment risk, supplier relations, and distribution of value over time.
  • Community-controlled evidence from cocoa farmers, farm workers, children, and supplier communities, alongside independent ecological evidence concerning land, forests, water, emissions, biodiversity, and future stewardship.
  • Comparative tests of whether different review forums improve correction, learning, representation, and remedy without becoming substitute governors.
  • Tests of all four Workloop proposals across routine, emergency, and unequal- power settings, including scapegoating, revoked access, chilled dissent, surveillance, and emergency authority that does not expire.

The Hershey dispute demonstrates that a formal power can face consequential review before it is finally exercised. It does not settle whose welfare should govern, which forum should decide, or whether the interrupted transaction would have served the institution over time.

Additional reciprocal institution comparisons

Newly developed institutional records add these reciprocal comparison paths:

Each path identifies a sourced case where this idea is a defining emphasis. The relation is editorial comparison, not evidence of direct influence, shared terminology, or equivalent outcomes.

Source notes

  1. Primary governing instrument: Milton Hershey School, Second Restated Deed of Trust (approved November 15, 1976), foreword and pp. 2–5, 8–10, on the 1909 parties and perpetual purpose, trustee and manager authority, investment approval, recordkeeping, vacancies, and succession, school-hosted deed. The restatement incorporates court-approved modifications through 1976; it is not a facsimile of each historical version, does not contain the 1918 stock transfer, and cannot establish later practice or participant experience.

    ↩ ↩ ↩ ↩

  2. Primary judicial record: In re Milton Hershey School Trust, 807 A.2d 324 (Pa. Cmwlth. 2002), “Appeal from Grant of Preliminary Injunction,” the attached September 10 Orphans' Court adjudication, and Judge Pellegrini's dissent, on the controlling interest, diversification rationale, bid process, attorney-general petition, potential public harm, injunction, and disagreement over the reviewing forum's authority, complete opinion. The court affirmed preliminary relief under a deferential reasonable-grounds standard; it did not decide final merits, prove every predicted harm, or preserve representative testimony from every affected group.

    ↩ ↩ ↩ ↩ ↩ ↩ ↩ ↩ ↩

  3. Independent legal history: Mark Sidel, “The Struggle for Hershey: Community Accountability and the Law in Modern American Philanthropy,” University of Pittsburgh Law Review 65 (2003), pp. 1–61, especially pp. 13–33 and 56–59 on community mobilization, the attorney general's changing position, litigation, termination of the sale process, and competing definitions of public interest, journal PDF. Sidel synthesizes filings and contemporaneous reporting close to the events and argues for a flexible community-accountability view; the article is not a participant-controlled archive or a representative welfare study.

    ↩

  4. Independent empirical counter-reading: Jonathan Klick and Robert H. Sitkoff, “Agency Costs, Charitable Trusts, and Corporate Control: Evidence from Hershey's Kiss-Off,” Columbia Law Review 108 (2008), p. 749 onward, especially the abstract, parts III–IV, and the sections on alternative hypotheses and the single-firm design, Harvard DASH manuscript. The event study estimates market-value reactions and advances an agency-cost interpretation. It does not directly measure beneficiary, employee, or community welfare, and the authors acknowledge single-firm and alternative- explanation limits.

    ↩

  5. Primary judicial record with a separate subject: In re Milton Hershey School and Hershey Trust Company, 911 A.2d 1258 (Pa. 2006), pp. 1–9, on the deed's board arrangement, the alumni association, the 2002 and 2003 administrative agreements, special-interest standing, and the attorney general's enforcement role, majority opinion. The holding concerns alumni standing to challenge modification of an administrative agreement, not the merits of the proposed company sale.

    ↩ ↩ ↩

  6. Primary theory: Michael C. Jensen and William H. Meckling, “Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure,” Journal of Financial Economics 3, no. 4 (1976), pp. 305–360, especially pp. 308–310 for the agency relationship and monitoring, bonding, and residual-loss components, article DOI. The economic model develops ownership and agency-cost propositions; it does not select Hershey's proper principal, document influence on the 2002 actors, or represent affected parties outside its contractual framing.

    ↩ ↩ ↩

  7. Primary theory: James H. Davis, F. David Schoorman, and Lex Donaldson, “Toward a Stewardship Theory of Management,” Academy of Management Review 22, no. 1 (1997), pp. 20–47, especially pp. 20–27 and 37–43 on agency and stewardship assumptions, psychological attributes, and situational characteristics, article DOI. The article proposes a model rather than measuring the Hershey trustees' motives or conduct, and organizational identification does not by itself establish inclusive or beneficial governance.

    ↩ ↩ ↩

  8. Primary conceptual framework: Mark Bovens, “Analysing and Assessing Accountability: A Conceptual Framework,” European Law Journal 13, no. 4 (2007), pp. 447–468, especially pp. 450–452 and 462–466 on the actor–forum relation and democratic, constitutional, and learning evaluations, article DOI. Bovens deliberately uses a narrow public-accountability concept; the framework did not shape the Hershey proceeding and does not decide which forum or affected parties deserve authority.

    ↩ ↩ ↩

  9. Current participant-controlled self-description: Milton Hershey School Alumni Association, homepage and mission statement, on fostering alumni community and supporting the principles of the deed, association site, accessed 14 July 2026. It establishes how the present association describes itself, not its full 2002 position, the views of all alumni, or the perspectives of current students and other affected groups.

    ↩

  10. Later corporate disclosure and independent sector research: The Hershey Company, 2025 Responsible Business Report, contents and “About This Report,” on the company's current self-selected sourcing, workforce, human-rights, and environmental reporting fields, company report. NORC at the University of Chicago, “Child Labor in Cocoa Production in Côte d'Ivoire & Ghana” (2020), on a U.S. Department of Labor-funded sector survey and hazardous work, research brief. The company document is a later management report, and NORC is sector-wide; neither attributes a measured supply-chain or ecological outcome to the trust's 2002 decision or substitutes for community-controlled evidence.

    ↩
Research record

Evidence basis

Claim Cited. Material claims carry source locators; comparative interpretation may still evolve.

Open questions and affected lives

Benefit-to-life status: Seed

  • Who may demand an account when legally authorized governors make a decision that reaches beyond the named beneficiary?
  • Which people receive standing through ownership, beneficiary status, public office, community dependence, or direct exposure to harm?
  • How can a review forum constrain governors without becoming an unaccountable substitute governor?
  • Who can represent workers, communities, nonhuman animals, ecosystems, and future generations when they cannot readily enter the governing forum?

These questions remain open; absence from the record does not imply absence of benefit or harm.

Ledger
I14
Research
Working
Evidence
Claim Cited
Ethical review
Seed
Structured atlas record

Lineage nodes

  1. The Hershey trust arrangementassigns school management and investment authority to interlocked trustee and manager offices serving a perpetual charitable purpose
  2. The proposed 2002 saleplaces portfolio diversification, beneficiary welfare, corporate control, and community consequences in conflict
  3. Public and judicial reviewrecords an attorney-general petition and preliminary injunction whose scope and legal basis were disputed
  4. Agency theorymodels delegation under divergent interests and positive monitoring, bonding, and residual costs
  5. Stewardship theorymodels conditions under which managers may identify with collective rather than solely individual objectives
  6. Accountability as an actor–forum relationshipdistinguishes a narrow process of explanation, questioning, judgment, and possible consequences from disclosure alone
  7. The contestable chain of governanceproposes an editorial diagnostic joining authorization, evidence, standing, review, remedy, and succession

Typed relationships

AThe Hershey trust arrangement → The proposed 2002 sale

Organizational Response: The trust instrument assigned investment decisions to the trustee with manager approval, and the trust's controlling voting interest enabled the 2002 sale process.

The restated deed and judicial record establish formal authority and control; they do not show that every affected group held beneficiary or enforcement rights.

AThe proposed 2002 sale → Public and judicial review

Organizational Response: The attorney general petitioned for review, the Orphans' Court enjoined any sale commitment, and the Commonwealth Court affirmed the preliminary restraint.

The appellate court reviewed whether reasonable grounds supported preliminary relief; it did not decide the ultimate merits or establish that every predicted harm would occur.

CAgency theory → The proposed 2002 sale

Interpretive Extension: Jensen and Meckling's agency-cost model raises a comparative question about delegated discretion, monitoring, and residual loss in the trust arrangement.

The model does not select the legally or ethically proper principal, and no evidence establishes that it shaped the trustees' decision.

CStewardship theory → The proposed 2002 sale

Parallel Solution: Stewardship theory raises a comparative question about when officeholders identify with an institution's objectives rather than pursue solely individual utility.

It supplies a behavioral model, not evidence of the Hershey trustees' motives, faithful conduct, or intellectual influence.

CAccountability as an actor–forum relationship → Public and judicial review

Interpretive Extension: Bovens's actor–forum framework distinguishes disclosure from a process that includes explanation, questioning, judgment, and possible consequences.

The framework does not determine which forum is legitimate or which affected parties deserve standing, and it did not shape the 2002 proceeding.

BPublic and judicial review → The contestable chain of governance

Empirical Reconstruction: The deed, sale dispute, preliminary review, and later standing decision can be organized as an editorial chain of authorization, evidence, standing, review, remedy, and succession.

The chain is not a holding of either court, a documented actor framework, or proof that the same sequence governs other institutions.

Provenance and sources

Corpus
management-research-library-2026-07-10
Source section
Governance, stewardship, and accountability
Imported
2026-07-10

Online anchors

Continue exploring

118 connected institutions