The Hearing / Sources & historical scope
Dated evidence, with its limits intact.
Country files are historical windows. They use different offices, periods, populations and measures. The original documents and supplied records support specific observations; a count, survey or media measure does not automatically identify a political effect.
Levada Center / 2 May 2024; survey fieldwork 21–28 February 2024, with a January hypothetical-ballot comparison
Find it: Opening and first three substantive paragraphs; Methodology
Source limit: Later publication reports attitudes within the historical window. Name recognition and hypothetical preferences cannot estimate support under unrestricted entry and campaigning.
Human Rights Watch / January 2024, reporting 2023
Find it: Counterterrorism and Counter-Extremism, first four paragraphs
Source limit: Rights-monitoring report documents organizational restrictions and characterizes prosecutions as politically motivated. It does not estimate electoral support or the votes those restrictions changed.
OSCE Office for Democratic Institutions and Human Rights and OSCE Parliamentary Assembly / 29 January 2024
Find it: Opening paragraphs and paragraph beginning “ODIHR engaged”
Source limit: Primary statement establishes that these bodies received no invitation and could not deploy. Non-observation does not measure fabricated votes or sincere support.
Levada Center / 24 May 2024; fieldwork 21–27 March 2024
Find it: Campaign ratings; Electoral preferences; Methodology
Source limit: In-person survey of 1,628 adults reports recalled votes and attitudes. Questions have different denominators. These are not official returns or estimates of preferences under open competition.
Human Rights Watch / January 2025, retrospectively reporting 2024
Find it: Freedom of Assembly, paragraph beginning “In March”
Source limit: HRW cites OVD-Info’s account of detentions. Before using a numerical total from the original reports, check whether it counts people or detention incidents; the units are not interchangeable. Neither source measures national electoral competitiveness.
Levada Center / 19 April 2024; fieldwork 21–28 February 2024
Find it: Opening methodological qualification; last substantive paragraph; Methodology
Source limit: A pre-election measure of sympathies, explicitly not a forecast. Restrictions already shaped choices; survey answers do not establish unrestricted competition.
Inter-Parliamentary Union / Election 5 May 2013; cites House of Representatives information dated 6 June 2013
Find it: Background; Distribution of seats; Sources
Source limit: Administrative seat totals and a compiled contemporary narrative. Opposition popular vote is approximately 51%, not a precise figure in this source. A vote-seat gap alone does not prove fraud.
Inter-Parliamentary Union / Election 9 May 2018; cites Parliament information dated 21 May and 17 July 2018
Find it: Background; notes 1–2; Distribution of seats; Sources
Source limit: PH total is 113 seats, excluding Warisan’s eight. Common ballot labels do not equal constituent-party membership. Narrative identifies plausible changes without isolating their causal effects.
Human Rights Watch / 2018, reporting 2017
Find it: Freedom of Expression; Freedom of Assembly and Association
Source limit: Rights-monitoring report documents restrictions, rather than estimating how many votes they changed or measuring the incumbent’s sincere support.
United States Department of Justice / 20 July 2016
Find it: Opening paragraphs and description of civil forfeiture complaints
Source limit: Announcement establishes enforcement action and allegations concerning 1MDB. It is not a conviction of Najib at that time or a measure of the controversy’s electoral effect.
Human Rights Watch / 2019, reporting 2018
Find it: Introduction; Freedom of Expression; Freedom of Assembly and Association
Source limit: Post-election evidence about institutional and rights constraints. It tests what turnover accomplished and cannot explain an election that had already occurred.
Inter-Parliamentary Union / Historical record of the system used for the 2013 and 2018 contests
Find it: Mode of designation; Constituencies; Voting system
Source limit: Documents 222 directly elected single-member constituencies and simple-majority voting. Other dated fields on this archived page should not be treated as current law.
European Union Election Observation Mission; document mirrored by EISA / 23 May 2019; media monitoring 17 April–18 May 2019
Find it: Printed pp. 1–2 and 9, especially footnote 27
Source limit: Preliminary statement issued before final tabulation and court petitions. Media percentages concern monitored party airtime, not audiences, beliefs or votes.
Malawi Electoral Commission / Report concerns May 2019, with narrative including the February 2020 judgment; current MEC file uploaded September 2026, original publication date not established
Find it: Printed p. 66, official candidate results; p. 72, petition and judgment account
Source limit: Declared 2019 results were later set aside. Other candidates totaled 290,446 of 5,031,264 valid votes (5.7728%, displayed as 5.8%). These administrative figures do not reconstruct an irregularity-free count. Current upload date is separate from the election date.
Malawi Electoral Commission / Report concerns June 2020; current MEC file uploaded September 2026, original publication date not established
Find it: Printed pp. 10–12, institutional steps; pp. 28–29, tickets and implementation; p. 41, counting; p. 42, final results
Source limit: Official retrospective administration account. Its candidate totals imply 59.33% for Chakwera and 39.93% for Mutharika. The separate MEC polling-station publication MW-E implies 59.34% and 39.92%. The reason for this difference is unestablished. Student values use one decimal, on which the series agree.
Mwiza Jo Nkhata, Anganile Willie Mwenifumbo and Alfred Majamanda / 2021, Journal of African Elections 20(2):57–80; DOI 10.20940/JAE/2021/v20i2a4
Find it: Printed pp. 59–60, court sequence; pp. 65–68, majority-rule reasoning
Source limit: Original scholarly analysis of the judgments, with criticism of parts of judicial reasoning. Original judgments could not be retrieved: court details rely on this analysis and MEC’s account, not a claimed direct reading.
Malawi Electoral Commission / Generated 29 July 2020; currently hosted file uploaded under a November 2022 path
Find it: p. 233, grand totals; p. 1, column headings
Source limit: Totals are Chakwera 2,604,043; Mutharika 1,751,877; Kuwani 32,456 (4,388,376 valid votes). MW-C instead reports 2,604,300; 1,752,530; 32,473 (4,389,303). The source discrepancy is distinct from the valid-vote/all-ballot denominator issue.
OSCE Office for Democratic Institutions and Human Rights / 28 February 2024; television monitoring 21 November–14 December 2023
Find it: Printed pp. 5–7, background and rules; pp. 13–14, campaign; pp. 20–21, media; pp. 27–28, challenges; pp. 32–33, REC final results
Source limit: Observer findings and official results reproduced from REC are distinct evidence. Recorded ballots do not independently validate free preferences. Media content does not measure audience exposure or persuasion.
Government of the Republic of Serbia / 30 October 2023
Find it: First three substantive paragraphs after the publication date
Source limit: Primary evidence of the government’s stated rationale, including its reference to opposition election demands. It does not establish equal campaigning conditions or verify its own democratic claims.
Center for Investigative Journalism of Serbia / 29 November 2023
Find it: Original recruitment narrative, recorded exchanges and cash-payment sections; exclude later related-story teasers
Source limit: Firsthand investigation, with recordings and separate source assertions, rather than a criminal conviction. One documented operation cannot establish how many national votes changed.
Levitsky, Steven, and Lucan A. Way. 2020. “The New Competitive Authoritarianism.” Journal of Democracy 31(1): 51-65. DOI. This conceptual reading is separate from the later case evidence and does not classify every case alike.