GOVERNING · Part II · The Captured Lottery
The numbers and sources
Chapter 9: The Corrupt and the Corrupted
The chapter tells its story with a handful of figures. Here is everything behind it: each figure with its study, its research document and its row in the log, the capture test and the scale ledger in full, the case for and the case against, where the sources disagree, and what is still open.
- 76 open
- 19 to check first
Numbers to carry
The figures the chapter is built on, each with its research document and its row in the log.
- A few in every hundred at most, perhaps nearer one: dark triad traits in strong measure in the population at large. CA-01 gives 1 to 4.5 per cent for psychopathy by threshold from a weak source and approximately 1 per cent in its own summary, and CA-02 says less than 1 per cent. Contested. CA-01, CA-02, C09-07. Owned here and cited by P-14 and C01-17.
- About four in every hundred: managers picked for executive development who scored at or above the clinical threshold on Hare's checklist. Babiak, Neumann and Hare 2010, Behavioral Sciences and the Law, N 203, 3.9 per cent. CA-01, C09-09. Other documents give 5.75 to 13.5 per cent (CA-02) and about 3.5 per cent (CA-06) for the same team's work.
- One chief executive in five: the 2016 claim from Bond University, retracted in 2018. Fallen. CA-01, C09-11.
- About six times in ten against four in ten: corrupt choices by laboratory volunteers given more power and fewer limits against those given less. Bendahan and colleagues 2015, Leadership Quarterly, 162 appointed leaders, 0.61 against 0.39. CA-02, C09-02.
- Nearly six households in ten against fewer than one in six: attendance at Indonesian relief meetings run by the whole community against those run by local leaders. NBER w18798, 59 against 15 per cent. CA-11, C09-28.
- A few points: the extra chance that village office holders and their kin received relief, compared with equally poor neighbours. NBER w18798, about 5, 8 and 3 points. CA-11, C09-29.
- About 150 people: where gossip stops keeping accurate accounts. PS-04 and CA-11, C09-50.
- A village of 300 and a district of 3,000: the rungs the In practice speaks to. Brief rung sizes, ME-14, CA-11, C09-63.
Every figure
One row for each figure or finding the chapter rests on. Status is how the finding stands in the research: robust, contested, open or fallen. It is not yet the log's verdict.
| Figure | What it measures | Source (authors, year, venue, N, or the record and its date) | Document | Row | Kind | Status |
|---|---|---|---|---|---|---|
| one to three followers | the size of each volunteer's charge in the game | Bendahan, Zehnder, Pralong and Antonakis 2015, Leadership Quarterly, 162 leaders | CA-02 | C09-01 | record | open |
| about six in ten, four in ten | corrupt choices with more and with less power | same, 0.61 against 0.39 | CA-02 | C09-02 | record | contested |
| 2002 | the dark triad named | Paulhus and Williams 2002 | CA-01 | C09-06 | record | robust |
| a few in every hundred, perhaps nearer one | population prevalence | 1 to 4.5 per cent by threshold from a vendor page, approximately 1 per cent, less than 1 per cent | CA-01, CA-02 | C09-07 | said | contested |
| about two hundred managers, 2010 | the sample | Babiak, Neumann and Hare 2010, Behavioral Sciences and the Law 28(2), N 203 | CA-01 | C09-08 | record | robust |
| about four in every hundred | at or above the clinical threshold | same, 3.9 per cent, PCL-R 30 or above | CA-01 | C09-09 | record | contested |
| several times | the executive rate against the population | CA-01 says approximately four times | CA-01 | C09-09 | inferred | contested |
| one in five, 2016, 2018 | the CEO psychopathy claim and its retraction | Brooks and Fritzon 2016, Bond University, retracted 2018 | CA-01 | C09-11 | said | fallen |
| modestly, almost nothing | narcissism and being chosen, and leading well | Grijalva et al. 2015, Personnel Psychology, rho 0.16, k 18, N 3,131, and rho 0.03, k 32, N 5,593 | CA-01 | C09-12 | record | contested |
| tens of thousands of people | the pooled triad sample | Landay, Harms and Credé 2019, Journal of Applied Psychology, k 46, N 32,680 | CA-01 | C09-13 | record | robust |
| higher | autocrats on psychopathy and narcissism | Nai and Toros 2020, Political Research Exchange, 124 candidates, 57 elections, 1,030 experts | CA-02 | C09-14 | record | robust |
| 2003 | approach and inhibition theory | Keltner, Gruenfeld and Anderson 2003, Psychological Review | CA-02 | C09-19 | record | contested |
| 1971 | the theory of regulatory capture | Stigler 1971 | CA-11 | C09-23 | said | contested |
| about two in five | retiring members of Congress who become lobbyists | Lessig 2011 via Wikipedia, 42 per cent | CA-11 | C09-25 | said | contested |
| billions | World Bank spending on community programmes | Mansuri and Rao 2013 | CA-11 | C09-27 | record | robust |
| nearly six in ten, fewer than one in six | attendance at relief meetings | NBER w18798, 59 and 15 per cent | CA-11 | C09-28 | record | robust |
| a few points | extra relief for office holders and kin | NBER w18798, 5, 8 and 3 points | CA-11 | C09-29 | record | robust |
| about 330 BC | the surviving antidosis speech | Demosthenes, Against Phaenippus | CA-11 | C09-31 | record | robust |
| 1096, half | the tithe and the chieftains' share | the Tithe Law of 1096 | PS-01, NO-07 | C09-32 | record | robust |
| 1220, a handful of families (five in PS-01, three to six across the documents, chapter 5 owns it at C05-61) | consolidation of the chieftaincies | the record of the Sturlung era | PS-01 | C09-33 | record | robust |
| 1973 and 1974 | Tanzanian villagisation | Scott 1998 via secondary sources | CA-12 | C09-36 | record | robust |
| June 1381 | the burning of court rolls | Walsingham via secondary chapters | ME-14 | C09-37 | record | contested |
| more than a millennium | the life of the Chinese censorate | secondary web sources | CA-12 | C09-39 | record | robust |
| 1625 to 1627, hundreds | the purge of the Donglin | Britannica and a Donglin study | CA-12 | C09-40 | record | contested |
| 1766 | the Swedish Freedom of the Press Act | Wikipedia, chydenius.net | CA-12 | C09-41 | record | robust |
| 1297, a year and a day | de Boochose's forged tally and sentence | Flegg via Scribd | ME-14 | C09-43 | record | contested |
| 1826, 1834 | the end of tallies and the Parliament fire | Fitznigel and secondary sources | CA-12, ME-14 | C09-44 | record | robust |
| two to four | knot-keepers per Andean community | Hyland 2022, Salomon 2004 | ME-14 | C09-46 | record | contested |
| about a third | criticism driven by envy round the fire | Wiessner 2014, 44 years among the Ju/'hoansi | PS-04 | C09-49 | record | robust |
| about 150 | the reach of gossip | Dunbar, Hutterite fission at 150 to 230 adults | PS-04, CA-11 | C09-50 | inferred | contested |
| a few hands | the Themistocles ostraka | via Wikipedia | PS-04 | C09-51 | record | contested |
| tens of thousands, most women | witch trial executions | Levack, Behringer, 45,000 to 50,000 executions, 75 to 85 per cent women, 1450 to 1750 | PS-04 | C09-52 | record | contested |
| several times larger | the power posing replication | Carney, Cuddy and Yap 2010, N 42, Ranehill et al. 2015, N 200 | CA-02 | C09-55 | record | robust |
| 2019 | the Stanford Prison Experiment taken apart | Le Texier 2019, American Psychologist | CA-02 | C09-56 | record | robust |
| 1940 | Churchill as the useful hard man | Van Vugt | CA-02 | C09-59 | inferred | contested |
| 300, 3,000, three | village, district, literate minimum | brief rung sizes and ME-14's design | ME-14, CA-11 | C09-63 | design | open |
The capture test
Leadership selection. Vector: campaigns and interviews reward confidence, charm and the performance of competence, which the checklist tracks. First move: win the room early. Precedent: Babiak's high potential managers. Resisting feature: the lot, and asking those who worked under a candidate. CA-01, CA-02.
Power once held. Vector: power with few limits, especially in people already high in testosterone. First move: take a little where nobody checks. Precedent: the laboratory game. Resisting feature: rotation and audit. CA-02.
The village purse. Vector: the literate volunteer who chairs the committee and sets meetings at harvest. First move: a quorum of tenants, debtors and cousins. Precedent: village councils before India's 1994 amendment. Resisting feature: posted accounts, full meetings, an outside check. CA-11.
Concealed wealth. Vector: invisible property that escapes public burdens and buys a hidden faction of debtors. Precedent: Athens, and Iceland after 1096. Resisting feature: the liturgy with the right of challenge, open stores. CA-11, PS-01.
The record. Vector: the clerk bribed or bullied into altering the land roll or the minutes, the debtor who destroys a tally, the strongman who seizes the archive. Precedent: the censorate under Wei Zhongxian, the monastic rolls altered after the plague. Resisting feature: split tallies, chirographs, two copies in two places, reading aloud. CA-12, ME-14.
Gossip. Vector: the manipulator carrying different stories to different circles. Precedent: the prepared ostraka, Salem, the witch trials. Resisting feature: first hand testimony, accusations heard in the open with the accused present. PS-04.
The scale ledger
| Arrangement | Household | Hamlet | Village | District | Watershed | Region | Nation | Planet | Document |
|---|---|---|---|---|---|---|---|---|---|
| Selection by lot against charm | untried | untried | bureaucratic | highly effective | optimal (Athens) | advisory only | advisory only | untried | CA-02 |
| Village elite capture of relief | not applicable | not applicable | frequent failure | mixed, high capture | mixed | central domain | central domain | not applicable | CA-11 |
| Lobbying and regulatory capture | not applicable | not applicable | never tried | moderate capture | frequent capture | high capture | maximum capture | high capture | CA-11 |
| Liturgy, lot and liability | not applicable | not applicable | functional | functional | highly functional | functional (vouchers) | never tried | never tried | CA-11 |
| Top down legibility | failed | failed | failed | operated | worked | worked | worked | untried | CA-12 |
| Censorates and social credit | failed | failed | failed | operated | worked | worked | worked | never tried | CA-12 |
| Analog records (tally, register, juries) | operated | worked | worked | worked | worked | operated | failed | failed | CA-12 |
| Bureaucratic writing and the cadastre | never tried | never tried | worked if locally held | worked | worked | worked | worked | failed | ME-14 |
| Organic gossip | works | highly effective | fails above about 150 | fails | fails | fails | fails | fails | PS-04 |
| Mob shaming | limited | limited | catastrophically effective | catastrophically effective | catastrophically effective | needs state media | needs state media | not given | PS-04 |
In a sentence: gossip and open records hold best from the household to the village, money captures personally in the village and district and through rules from the town upward, and legibility and watchers serve whoever holds them from the district to the nation.
The case for and the case against
The case the chapter makes
Both stories about power are true. People high in the dark triad are a small minority, but they are chosen more often than they deserve because their confidence looks like competence, and ordinary people given power with few limits become more willing to take. The commonest routes to capture are quiet ones, money and records, and they are opened by people who look helpful. The defences are mostly bookkeeping: nobody handles the purse alone, records are kept in two places, read aloud and readable by those they concern, and talk may start an inquiry but not finish one. CA-01, CA-02, PS-01, CA-11, CA-12, ME-14, PS-04.
The case against
The laboratory evidence on power has a poor replication record (power posing, the Stanford Prison Experiment, a hypocrisy study coauthored by a psychologist later found to have invented data). The dark triad may be a renamed mix of low agreeableness and low conscientiousness (Miller and Lynam). The surviving effects are small. A hard leader can save a group in crisis (Van Vugt, Churchill), guarding every door produces vetocracy (Fukuyama), some elite capture is pro poor (Dasgupta and Beard), and legibility also brought material gains (Tauger against Scott). The chapter's reply is that its rules do not depend on diagnosing anyone. CA-01, CA-02, CA-11, CA-12.
Where the sources disagree
- Prevalence. CA-01 gives 1 to 4.5 per cent by threshold from a vendor page and approximately 1 per cent in its summary. CA-02 says less than 1 per cent. For executives CA-01 gives 3.9 per cent of 203 (closest to the paper), CA-02 gives 5.75 to 13.5 per cent for the same paper via Semantic Scholar, and CA-06 gives about 3.5 per cent from Snakes in Suits via Goodreads. The chapter carries a few in every hundred and about four in every hundred. ⚠
- Grijalva et al. 2015: the printed interval of .08 to .15 does not contain the estimate of 0.16. ⚠
- Bendahan and colleagues: a single laboratory study, and CA-02 describes the contrast behind 0.61 and 0.39 as both power and constraint. ⚠
- Power and hypocrisy: CA-02 relies on Lammers, Stapel and Galinsky 2010 without noting the Stapel fraud case, cites Kipnis 1972 to an unrelated EconStor paper, and cites its replication figures to a paper on anchoring. None of these figures is printed. ⚠
- Priming and power: the card game findings and the threat findings come from a priming literature with a weak replication record. ⚠
- The 42 per cent of retiring members of Congress who lobby reaches CA-11 through Wikipedia. ⚠
- Witch trial figures reach PS-04 through a secondary chain. ⚠
- Domesday: CA-12 treats its public juries as a defence against forgery, ME-14 as a conqueror's instrument of total surveillance. Left out.
- Scott against Tauger on whether legibility destroys or protects carrying capacity. CA-12.
Open questions
- No research document describes the biscuit experiment. The Preface no longer promises it (P-40 reworded 2026-10-23 to the money game, C09-01 and C09-02). A source would be needed before it could return.
- Is the dark triad one thing, or three, or a renaming of ordinary traits?
- Did the Indonesian capture matter for the welfare of the poor, or only for the lists?
- How can a mostly illiterate village read its own register? ME-14 proposes weekly public recitation and split tallies, untested after a collapse.
- Where exactly does gossip stop working, at about 150 or nearer 230?
- Can a village of 300 keep a ballot secret when everyone knows everyone's hand? ME-14 leaves it open.
- CA-07 and CA-08 were never delivered, so the capture domain is thinner than planned.
The documents
Babiak, P., Neumann, C. S. and Hare, R. D. (2010). Corporate psychopathy: talking the walk. Behavioral Sciences and the Law 28(2). Bendahan, S., Zehnder, C., Pralong, F. P. and Antonakis, J. (2015). Leader corruption depends on power and testosterone. The Leadership Quarterly. Grijalva, E. et al. (2015). Narcissism and leadership. Personnel Psychology 68(1). Landay, K., Harms, P. D. and Credé, M. (2019). Shall we serve the dark lords? Journal of Applied Psychology 104(1). Keltner, D., Gruenfeld, D. H. and Anderson, C. (2003). Power, approach, and inhibition. Psychological Review. Mehta, P. H. and Josephs, R. A. (2010). Testosterone and cortisol jointly regulate dominance. Hormones and Behavior. Nai, A. and Toros, E. (2020). The peculiar personality of strongmen. Political Research Exchange. Henrich, J. and Gil-White, F. J. (2001). The evolution of prestige. Evolution and Human Behavior. Ranehill, E. et al. (2015). Assessing the robustness of power posing. Psychological Science. Le Texier, T. (2019). Debunking the Stanford Prison Experiment. American Psychologist. Stigler, G. (1971). The theory of economic regulation. Mansuri, G. and Rao, V. (2013). Localizing Development: Does Participation Work? World Bank. NBER working paper w18798. Does Elite Capture Matter? Local Elites and Targeted Welfare. Scott, J. C. (1998). Seeing Like a State. Yale University Press. Fitznigel, R. (about 1179). Dialogus de Scaccario. Salomon, F. (2004). The Cord Keepers. Duke University Press. Feinberg, M., Willer, R. and Schultz, M. (2014). Gossip and ostracism promote cooperation in groups. Psychological Science. Levack, B. P. (1987). The Witch Hunt in Early Modern Europe. Nussbaum, M. C. (2004). Hiding from Humanity. Princeton University Press.