← Atlas
Idea

Measurement, accounting, and control

A measure redistributes authority: someone defines the category, someone produces the record, someone acts on it, and someone bears what the account excludes. Colombia's false-positives system shows how a performance proxy can organize work around producing the record rather than the reality it was meant to describe, while families and investigators show how counter-evidence can reopen a result to contradiction and responsibility.

Governing questionWhat happens when a measure of success becomes valuable enough to reorganize the work it was supposed to observe?

Working · Claim Cited

A reported death could be two incompatible records

Between January and August 2008, nineteen young people from Soacha and Bogotá disappeared. Their families later learned that bodies had been found hundreds of kilometres away in Ocaña and Cimitarra and presented as guerrillas killed in combat. The families' record began with a missing person, a name, a relationship, and a search. The military record ended with an operational result.1

That contradiction was not a harmless classification error. In Colombia's military false-positives system, civilians were murdered and represented as combat deaths. Recruiters sometimes used promises of work to move people away from home; soldiers killed them, altered scenes with weapons or clothing, and reported the deaths as lawful combat outcomes. Families searched while some victims were buried unidentified.2

No metric committed a murder. People exercised criminal agency, participants held unequal knowledge and authority, and responsibility must be determined for particular conduct. The measurement claim is narrower: when reported combat deaths became a valued sign of success, the count helped organize pressure, reward, comparison, documentation, and review around a result that could be falsified.

A proxy gained authority without becoming military effectiveness

Operational effectiveness includes lawful protection of civilians, territorial control, intelligence quality, force protection, captures and demobilizations, and the weakening of armed groups. A combat-death count observes none of those outcomes by itself. It is legible to a headquarters, however: units can report it, commanders can compare it, and senior officials can aggregate it.

The Special Jurisdiction for Peace (JEP) identifies pressure to present combat deaths, competition among units, incentives for operational results, deficient control and investigation, dismissal of complaints, and stigmatization of regional populations as factors in macrocriminal patterns. Its current Case 03 profile describes 6,402 people as the preliminary universe for 2002–2008 after comparison of participant accounts with prosecutorial, disciplinary, criminal, memory, and civil-society records. The number is period-bounded and provisional; the same profile reports other datasets with different totals and time windows.2

The sources do not warrant one simple national-policy claim. In 2010, United Nations special rapporteur Philip Alston found a repeated nationwide pattern too large to dismiss as isolated misconduct, but no evidence that senior government officials had ordered the killings or that they formed an official national policy. Later JEP findings state that, in two specified subcases, the crimes would not have occurred without an institutional body-count policy, incentives, and constant command pressure. The propositions differ in date, evidentiary record, institutional author, and scope; neither should be silently substituted for the other.2

Independent econometric research finds substantially more false positives during the high-powered-incentive period in municipalities with weaker judicial institutions and a larger share of brigades commanded by colonels, alongside no discernible improvement in overall security in the latter comparison. The authors explicitly caution that their estimates do not identify causal effects and cannot eliminate all time-varying alternatives. That evidence strengthens the incentive-and-weak-review mechanism without proving that one incentive, rank, or metric caused any individual killing.2

The result required a production and validation process

The documented pattern joined distinct acts. A recruiter or informer identified or lured a person. Transport separated the person from family and familiar institutions. Military personnel killed the victim. Weapons, uniforms, or other scene elements made the death resemble combat. The death was then reported as an enemy killed in action. JEP's synthesis describes planning, execution, and concealment, including deaths reported in documents available to superiors.2

The division of work could separate each participant from parts of the whole. A recruiter did not need to write the operational account. A person reviewing a result table did not necessarily encounter the family searching elsewhere. A superior could have a reputational stake in the unit's total. A soldier who objected faced a hierarchy in which Alston found it difficult to speak out. Civilian and military justice systems disputed jurisdiction, delayed cases, and held unequal access to evidence.2

The administrative record mattered because it made the local act portable. A body accompanied by a combat narrative could travel upward as comparable performance. The murdered person's history, the family's search, and doubts about the scene remained distributed across other people, places, and files. The result channel was integrated; the contradiction channel was not.

That asymmetry is an instance of organizational ignorance. It does not imply that every superior knew the truth or that ignorance was uniform. It asks which facts were routinely joined before a result conferred recognition, leave, promotion, money, or institutional standing, and which facts required an extraordinary effort to become audible.

Families made contradiction governable

MAFAPO's own account describes families discovering the distant burial and combat classification of nineteen young people, organizing despite threats, and submitting a victim-authored report to the JEP. The collective's account is not an independent national enumeration. It is evidence of what the families say they experienced, how they organized, and what truth and justice they continued to demand.1

At an October 2019 public hearing, thirteen family members responded to thirty-one voluntary accounts from members of two military units. The JEP described them as having been forced to become investigators of their own cases and treated their observations as a threshold against which contributions to truth had to be assessed. The hearing record covers a specified proceeding and does not stand for every victim or reproduce every testimony in full.1

The families' contribution was not decorative context around an official number. Missing-person histories, names, dates, relationships, and searches challenged the category that had made a civilian into an enemy result. Investigators could then compare those accounts with forensic evidence, operational records, testimony, unit patterns, rewards, and command conduct. Counter-accounting changed both the count and the questions: who recruited, who killed, who staged, who reported, who approved, who rewarded, who investigated, and who made contradiction costly?

Accountability remains unfinished. The national universe is provisional; different participants held different knowledge and authority; and truth, sanction, reparation, identification of remains, and guarantees of non-repetition cannot be reduced to one completion score. A throughput target for confessions or case closures could itself displace victims' purposes if the administrative result became more important than the truth and repair it was meant to serve.

Measurement traditions allocate knowledge and decision rights differently

The Colombian case is neither the origin nor the inevitable endpoint of organizational measurement. Four comparisons isolate different authority arrangements without claiming historical influence on Case 03.

At DuPont and General Motors under Alfred Sloan, Donaldson Brown's return-on-investment measure, forecasting practices, flexible budgets, and reporting supported comparison and coordinated control across decentralized businesses. A common financial account helped a central office allocate capital while operating divisions retained authority. The same account could not represent every worker, customer, community, or ecological effect unless those effects entered its definitions.3

Frederick Taylor's The Principles of Scientific Management assigned management the work of gathering, classifying, and formalizing craft knowledge. The planning room specified the task, method, and time, while the worker executed written instructions. Taylor presented this as shared responsibility and cooperation; it nevertheless moved authority over method toward management and its records.4

W. Edwards Deming distinguished special causes that may be local from common causes that belong to the system and require management action. His statistical argument was not simply “measure more.” It asked whether figures distinguish the level at which useful action is possible and whether an information system guides action rather than merely producing volume. Out of the Crisis placed that argument inside a broader transformation of management.5

Donald Campbell argued that a quantitative social indicator becomes more vulnerable to corruption and more likely to distort the process it monitors as it carries more weight in social decisions. He described this as a pessimistic, U.S.-bounded proposition and characterized much of his supporting evidence as anecdotal. It is a warning mechanism, not a universal law that proves the causes of the Colombian crimes.6

The comparison is therefore about decision rights, not resemblance alone. DuPont and GM used financial measures for delegated capital control; Taylor used measured knowledge to plan tasks; Deming used repeated data to distinguish responsibility for variation; Campbell warned about high-stakes indicators. None of those sources establishes that Colombian actors adopted the writers' ideas or corporate practices.

A measure needs a contestable constitution

“Measurement constitution” is a proposed design rule, not a JEP finding or a validated cross-case theory. It asks that a consequential measure disclose at least nine things:

  1. the purpose it is meant to serve;
  2. the object and operational definition being counted;
  3. who produces and preserves the underlying record;
  4. what independent evidence can contradict that record;
  5. which decisions, rewards, or sanctions the result authorizes;
  6. how uncertainty and ordinary variation are represented;
  7. which side effects and excluded values are monitored;
  8. how affected people can inspect, challenge, and correct the classification; and
  9. who can revise or retire the measure, under what evidence and review date.

Precision does not answer those questions. A precisely computed proxy can still misstate its purpose, concentrate proof in the judged unit, reward production of the record, or exclude the people who bear its consequences. The governing test is relational: who defines, who observes, who decides, who may contradict, and who bears what remains outside the account?

The boundary must extend beyond the immediate organization. Financial return can omit worker health and community costs. Output can omit sleep, housing, and unpaid care. Conservation counts can move harm outside a measured area. Digital measurement can add surveillance, energy, water, hardware, and supply-chain effects. No source reviewed here shows that the nine-part proposal prevents those omissions; each must be tested in the setting where it will govern.

Four product hypotheses remain unvalidated

The four Workloop proposals are research hypotheses, not demonstrated remedies:

  • I07-P01 proposes metadata for purpose, owner, definition, decision use, failure modes, and review date. Evaluation must test whether disputed definitions become resolvable and measures are revised or retired, while checking whether metadata merely adds ceremony or expands invasive collection.
  • I07-P02 proposes pairing process indicators with external result indicators. Tests must examine predictive validity, decision quality, and documented tradeoffs, including whether people simply learn to game both proxies.
  • I07-P03 proposes variation-aware review where repeated-process data meet the statistical assumptions. Tests must compare false alarms, detection time, recovery, and improvement quality while checking whether a poor model normalizes structural harm or suppresses a novel warning.
  • I07-P04 proposes scheduled proxy audits. Evaluation must show whether audits find drift or gaming early and whether revision improves the decisions served, while checking for ritual compliance, hidden targets, and exclusion of people subject to the measure.

The Colombia record, accounting history, and measurement theories motivate the questions. They do not evaluate any of the four mechanisms in a Workloop deployment.

Seven nodes and six relations bound the structured argument

The seven nodes move from two incompatible records to a proposed governance rule:

  • I07-N01 preserves families' missing-person records and life histories;
  • I07-N02 identifies reported combat deaths as the result proxy;
  • I07-N03 joins recruitment, killing, staging, reporting, and validation as a criminal production process;
  • I07-N04 contrasts aggregate success with distributed contradiction;
  • I07-N05 joins family testimony and investigative records as a counter-account;
  • I07-N06 compares delegated control, task standardization, system learning, and high-stakes social judgment; and
  • I07-N07 states the proposed, still-untested constitution of a measure.

The first five nodes are bounded by the Colombia and family records; I07-N06 is bounded by the four comparative traditions; I07-N07 is a design proposition, not an empirical result.213456

The six relations carry different evidence strength:

  • I07-E01, grade A, relates the combat-death proxy to the criminal process. JEP, UN, and independent research support pressure, competition, incentives, repeated methods, and deficient review. The qualification denies metric-only causation and preserves individual agency.
  • I07-E02, grade A, relates the criminal process to aggregated success. JEP supports planning, concealment, operational reporting, and documents known to superiors. The qualification denies one command, identical knowledge, or one workflow in every case.
  • I07-E03, grade A, relates family records to the counter-account. MAFAPO and the JEP hearing support family search, reporting, observation, and testimony; the qualification credits journalists, advocates, prosecutors, courts, the Truth Commission, and the JEP rather than making families act alone.
  • I07-E04, grade A, relates distributed contradiction to empirical reconstruction. The joined records support the gap between reported combat performance and civilian lives; the qualification keeps the victim universe and adjudication open.
  • I07-E05, grade C, compares Taylor, DuPont, GM, Deming, and Campbell as different allocations of knowledge and decision authority. Its five internal source IDs support endpoints, not influence on Colombian actors or a common lineage.
  • I07-E06, grade C and still working, extends the counter-account into the proposed contestability rule. The Colombia source ID supports the case prompt, not the rule's general validity, and the proposition is not a holding of the JEP.

Evidence grades describe support for each bounded relation, not importance, effect size, or moral severity. I07-E01 through I07-E04 use the Colombia institution record as their internal evidence route. I07-E05 uses the DuPont, GM, Deming, Out of the Crisis, and Principles of Scientific Management records. I07-E06 uses the Colombia record only for its empirical endpoint.

Seven related records are routes, not prerequisites

The untyped related_ids field contains seven records:

No typed idea_ids or work_ids are recorded, and no structured reading dependency is assigned. The routes can be read in any order; relatedness does not establish agreement, influence, or shared moral standing.

Established institutions are comparison paths, not one class

The institution coordinates are prompts for testing how organizations define, produce, aggregate, contest, and act on measures. Their own records supply their evidence. Listing them together asserts no common metric, result, causal lineage, performance level, or moral equivalence.

Historical, industrial, financial, and operational-control paths:

Corporate ownership, strategy, growth, decline, and harm paths:

Community, cooperative, service, and shared-resource paths:

State, coercive, extractive, industrial-labor, and cross-border paths:

Human consequences are documented; net impact is unclassified

No structured impact record is assigned. That absence does not mean no impact. The Colombia sources document murders, forced disappearances, altered burial and identity records, threats, delayed justice, family search labor, and continuing demands for truth and repair.21

The affected-party path includes murdered civilians; families and communities; soldiers and civilians with different roles and degrees of coercion, knowledge, and agency; whistleblowers; investigators; prosecutors; judges; journalists; advocates; the armed forces; and the public whose security claims depended on the result. The reviewed sources do not support one net classification across those parties, all regions, or the full period of the crimes.

No reviewed source measures ecological effects, future-generation effects, or the resource costs of the record, investigation, exhumation, adjudication, and reparation systems. The institution links include cases where land, water, animals, extraction, or future generations are central, but the links add no evidence about those effects here. Impact therefore remains unclassified rather than beneficial, harmful, neutral, or absent as one aggregate.

Evidence still needed

  • Direct citation to the underlying JEP orders for every legal proposition now supported by the official Case 03 summaries, with paragraph-level comparison of accepted findings, contested responsibility, and later procedural change.
  • A reconciled, versioned victim dataset that explains inclusion rules, duplicate handling, geographic and temporal scope, missingness, and changes from the provisional 6,402 figure.
  • Records from a wider range of regions, ranks, units, defendants, dissenting soldiers, investigators, and civilian officials to distinguish shared pattern from local variation and individual responsibility.
  • Community-controlled testimony and oral history beyond the Soacha proceedings, including families who have not participated in public hearings and people exposed to retaliation or unresolved identification.
  • Comparative studies that test whether independent evidence, appeal rights, reward separation, and revision authority actually reduce proxy corruption across military, corporate, public-service, and community settings.
  • Independent evidence on how Taylorist standards, ROI systems, and variation-aware management affected workers, customers, communities, and ecosystems across implementations rather than in programmatic or executive accounts alone.
  • Prospective and retrospective trials of I07-P01 through I07-P04, including failed deployments, false reassurance, gaming, burden on affected people, privacy, labor displacement, and accessibility.
  • Lifecycle evidence for the energy, water, hardware, land, and supply-chain effects of contemporary measurement systems, allocated to actual designs rather than assumed from aggregate digital infrastructure.

Paths into deeper study

  • Read the JEP pattern findings beside the MAFAPO and family-hearing records; ask what each source can establish and which authority or experience it cannot represent.
  • Compare ROI, Taylor's task and time standard, Deming's control chart, and the combat-death count by asking which decision each authorizes and whose knowledge can interrupt it.
  • Continue into organizational ignorance for the production of unavailable contradiction and into Colombia's military false-positives system for the institutional case.
  • For one consequential live metric, trace its definition, source record, reward or sanction, variation model, affected parties, side effects, appeal route, and conditions for revision or retirement.

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. Victim-authored memory and an official hearing summary. MAFAPO, “Una década sin respuesta para las madres de Soacha” (October 10, 2018), identifies MAFAPO as author and photographer and describes the nineteen disappearances, distant recovery and combat classification, organizing under threat, and submission of its report, Centro Nacional de Memoria Histórica. The source preserves the collective's voice and claims; it is an advocacy and memory account hosted by a state institution, not an independent national count or adjudication. “13 familiares de las víctimas de Soacha tuvieron la palabra en la JEP” (October 17, 2019) records the participants, thirty-one accounts under review, MAFAPO's earlier report, and the magistrate's description of families becoming investigators, Jurisdicción Especial para la Paz. This is an official institutional summary of a specified hearing; it does not reproduce all testimony, represent all families, or independently judge every assertion made there.

    ↩ ↩ ↩ ↩ ↩

  2. Official adjudicatory synthesis, independent international investigation, and peer-reviewed quantitative analysis. JEP's Archivos Vivos Case 03 overview identifies the six pattern factors and the planning, execution, and concealment finding under the closing “Caso 03” overview; accessed July 14, 2026, Jurisdicción Especial para la Paz. The current Case 03 profile gives the preliminary 6,402 figure, identifies the records compared, reports alternative datasets and periods, and limits the institutional-policy statement to two subcases; “¿En qué va el Caso 03?” and “Perfil del Caso 03,” accessed July 14, 2026, Jurisdicción Especial para la Paz. Both pages are official, mutable summaries of an ongoing jurisdiction; they are not independent reviews or substitutes for the underlying orders. Philip Alston, Report of the Special Rapporteur on Extrajudicial, Summary or Arbitrary Executions: Mission to Colombia, A/HRC/14/24/Add.2 (2010), paras. 10–15, 19–29, 37–41, and 89–95, documents the repeated pattern, recruitment and staging, disputes over policy and count, pressure, incentives, weak accountability, jurisdictional barriers, and recommendations, United Nations. The UN report supplies independent international fact-finding based on a 2009 mission; it predates later JEP evidence and cannot supply a current victim count or resolve later individual cases. Daron Acemoglu, Leopoldo Fergusson, James A. Robinson, Dario Romero, and Juan F. Vargas, “The Perils of High-Powered Incentives: Evidence from Colombia's False Positives,” American Economic Journal: Economic Policy 12, no. 3 (2020), pp. 1–13 and 35–38, American Economic Association. This is independent peer-reviewed municipality-level analysis with explicit data and identification limits; the authors state that their estimates do not identify causal effects and cannot rule out all time-varying factors, so it supports association and mechanism rather than individual culpability.

    ↩ ↩ ↩ ↩ ↩ ↩ ↩ ↩

  3. Dale L. Flesher and Gary John Previts, “Donaldson Brown (1885–1965): The Power of an Individual and His Ideas over Time,” Accounting Historians Journal 40, no. 1 (2013), abstract and article, University of Mississippi eGrove. The independent accounting-history study supports Brown's 1914 expanded ROI measure, forecasting and planning, decentralized corporate management, and early-1920s GM flexible budgeting. Its focus on Brown and financial innovation does not establish every operating effect or represent workers, customers, communities, and ecological consequences.

    ↩ ↩

  4. Frederick Winslow Taylor, The Principles of Scientific Management (1911), chapter II, especially the passages beginning with management's duty to gather and formalize worker knowledge, the planning room, the task idea, and the warning that time study can be used as a club, Project Gutenberg. This is Taylor's primary programmatic account. It establishes what he advocated and acknowledged, not independent evidence of implementation, worker consent, health, productivity, or long-run effects; pagination varies by digital format, so section and passage locators are supplied.

    ↩ ↩

  5. Out of the Crisis, 2018 reissue page, including bibliographic data, the 1982 original-publication note, and description of the fourteen-point management argument, MIT Press. The publisher page is an authoritative bibliographic destination and summary, not an independent test of results. W. Edwards Deming, “Some Statistical Logic in the Management of Quality” (1971), pp. 1–8 and 12–13, distinguishes special from common causes, locates responsibility for system causes in management, and argues that information systems should guide action, W. Edwards Deming Institute. The preserved paper is Deming's primary argument in a manufacturing-quality context. Its strong effectiveness claims are not independent estimates, and transfer to military, service, or software settings requires separate evidence.

    ↩ ↩

  6. Donald T. Campbell, “Assessing the Impact of Planned Social Change,” 1976 paper reprinted in Journal of MultiDisciplinary Evaluation 7, no. 15 (2011), printed pp. 34–35 under “Corrupting Effect of Quantitative Indicators,” Journal of MultiDisciplinary Evaluation. This is Campbell's primary conceptual and evaluative argument. He expressly bounded the proposition to the U.S. setting and called the evidence offered there predominantly anecdotal; it motivates a corruption-risk hypothesis but does not establish a universal law or the causes of Case 03.

    ↩ ↩
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 defines a measure, produces its evidence, and gains authority to act on the result?
  • Whose work, wellbeing, life, or ecological contribution remains outside the operational definition?
  • Can the people classified by a record inspect, challenge, and correct it before rewards or punishments follow?
  • Which independent evidence can reveal that a measure is corrupting the process it was intended to monitor?

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

Ledger
I07
Research
Working
Evidence
Claim Cited
Ethical review
Seed
Structured atlas record

Lineage nodes

  1. The families' missing-person recordspreserve names and life histories that contradict official descriptions of combat deaths
  2. Combat deaths as a result measureturn a count of reported enemy deaths into evidence of operational success
  3. A criminal production processjoins recruitment, murder, staging, paperwork, and validation to produce credible reported results
  4. Aggregated success and scattered contradictionmoves comparable totals upward while leaving disconfirming facts across families, graves, jurisdictions, and files
  5. The counter-accountjoins family testimony, missing-person reports, operational records, and investigations to restore names to the account and reopen responsibility
  6. Competing purposes of measurementcontrasts measurement for delegated control, work standardization, system learning, and high-stakes social judgment
  7. The constitution of a measuremakes definition, observer independence, decision rights, side effects, appeal, and revision part of the control system

Typed relationships

ACombat deaths as a result measure → A criminal production process

Operationalization: Pressure, competition, incentives, and weak review made reported combat deaths valuable enough to support repeatable criminal routines.

No measure caused a murder; people exercised criminal agency within an arrangement that made a validated death operationally useful.

AA criminal production process → Aggregated success and scattered contradiction

Organizational Response: Falsified operational records allowed local crimes to travel upward as comparable evidence of military performance.

The pattern does not establish that every case followed one command or that all officials possessed the same knowledge.

AThe families' missing-person records → The counter-account

Power And Voice Extension: Families' sustained searches, reports, and testimony supplied contradictions that the JEP and other institutions compared with military accounts and records.

Families did not act alone; journalists, advocates, prosecutors, courts, the Truth Commission, and the JEP supplied additional forums and evidence.

AAggregated success and scattered contradiction → The counter-account

Empirical Reconstruction: Joined-up testimony and records reconstructed the gap between reported combat performance and murdered civilian lives.

The JEP's victim universe remains provisional and accountability for different ranks and cases remains under adjudication.

CCompeting purposes of measurement → Combat deaths as a result measure

Interpretive Extension: Taylor, DuPont and GM, Deming, and Campbell expose different ways a measure can allocate knowledge and decision authority.

These are analytical comparisons; there is no evidence that the Colombian actors derived their result system from these writers or companies.

CThe counter-account → The constitution of a measure

Interpretive Extension: The counter-account motivates a proposed rule that a measure cannot support legitimate control when the result producer monopolizes its initial proof and affected people lack a route to challenge it.

The Colombian case supports the empirical prompt, not the rule's general validity; the proposed measurement constitution is neither a JEP holding nor a tested cross-case result.

Provenance and sources

Corpus
management-research-library-2026-07-10
Source section
Measurement, accounting, and control
Imported
2026-07-10

Online anchors

Continue exploring

73 connected institutions