Methodology

By Gil Guerra · Updated September 2026 · Evidence tiers

01

The Murillo broadcast corpus

Rosario Murillo, now co-president, speaks on state media almost every weekday at noon, and regime outlets publish transcripts of these routine broadcasts. The corpus used on this site contains 2,405 of those transcripts, from April 2012 through September 4, 2026: about 4.8 million words in 100,328 paragraphs.

Figure 1

Since 2016 the corpus holds roughly 240–260 transcripts in most full years, with a 12-month gap in 2019–20.

Routine Murillo broadcast transcripts per calendar year, by publishing outlet. August 2015 to September 4, 2026 (2,404 transcripts).

  • Canal 4
  • Radio Nicaragua
  • El 19 Digital

Source: Canal 4, Radio Nicaragua and El 19 Digital transcripts; author’s corpus.

How to read: each column is one year; its height is the number of daily transcripts collected, and the colors show which outlet published them. A full year of weekday broadcasts is roughly 250, so a shorter column means transcripts are missing or fewer broadcasts took place; the corpus cannot tell which.

How this was calculated

  1. Data: /data/home/corpus_coverage.json, built from the corpus index (one row per transcript, with its broadcast date and the site that published it).
  2. Each transcript is counted in the calendar year of its broadcast date and in the outlet of its publishing site: canal4.com.ni (Canal 4), radionicaragua.com.ni (Radio Nicaragua), el19digital.com (El 19 Digital).
  3. Column height = transcripts in that year; segments = transcripts per outlet; the number above each column is the total. Where two outlets published the same broadcast, only one copy is in the corpus.
  4. Years shown: 2015 (from Aug. 6, 2015, the first regular transcript) to 2026 (through Sept. 4, 2026). The one earlier transcript is in the corpus total (2,405) but not plotted.
  5. Shaded band: the 12 calendar months since August 2015 with no transcripts (Sept. 2019 to Aug. 2020), from the corpus’s monthly coverage table.

Notes and caveats

2015 starts in August; 2019 ends in August and 2020 starts in September; 2026 runs to September 4. One transcript from 2012 is in the corpus but not shown.

No transcripts were collected from September 2019 through August 2020 (shaded). Missing days are unknown, not zero, and are excluded from every count on this site.

Where two outlets published the same broadcast, only one copy is kept.

Sources

Transcripts published by Canal 4, Radio Nicaragua and El 19 Digital; corpus assembled by the author. See section 1.

YearCanal 4Radio NicaraguaEl 19 DigitalTotal
201510200102
201624300243
201725900259
201824400244
2019890089
2020880088
202123600236
2022881131202
20236106142254
202400253253
202500255255
202600179179
  • Aug. 2015effective start of coverage
  • 12 monthswith no transcripts, Sept. 2019–Aug. 2020
  • 3 outletsCanal 4, Radio Nicaragua, El 19 Digital
  • Checked by the authordictionary search plus AI-assisted reading; minor errors may remain

Figure 1 above shows how many transcripts the corpus holds for each year, by outlet.

How the measures are made

Counts come from dictionary search (accent-insensitive keyword and phrase patterns, listed in section 2), followed by AI-assisted reading of the matched sentences to classify how a topic is mentioned, for example Russia as a security partner versus Russia as a cultural or tourism partner. These counts were checked by the author, but minor errors may remain. They describe her routine daily broadcasts only, not speeches at special events, other officials’ statements or everything the state does. That is why findings on this site are worded “in her daily broadcasts” or “by this count,” and never as claims about what the regime as a whole has or has not said.

Learn more: coverage, outlets, voice and quotations

Coverage. Only one transcript survives from before August 2015, so effective coverage starts in August 2015. Since then, the only calendar months with no transcripts are the 12 from September 2019 through August 2020. Weekends and missing days are unknown, not zero: a count of “no mentions” never includes them.

Outlets. The source outlet changes over time: Canal 4 (2015–2023), Radio Nicaragua (2022–2023) and El 19 Digital (2023–2026). Canal 4 transcripts before 2022 are roughly twice as long as later ones, so shares of broadcasts that mention a term are not strictly comparable across the switch. Where two outlets published the same broadcast, only one copy is kept. Standard boilerplate (share prompts, sign-offs, headings) is masked before counting.

Voice. Murillo often reads government communiqués aloud. The corpus does not separate her own words from statements she reads, so figures describe what was said on her broadcast, not what she personally wrote.

Quotations. Every quotation from Murillo shown on this site is an exact match to the corpus text, carries the broadcast date and links to the original published page. English glosses are the author’s.

02

Topic dictionaries

Each topic on the Murillo’s Broadcasts page is defined by a list of search patterns and exclusions. A broadcast counts toward a topic if at least one pattern matches after exclusions are removed.

Figure 2

Spot checks found 94–100% of each dictionary’s sampled matches on topic.

Broadcasts matched by each of the 15 topic dictionaries, of 2,404 in the corpus (Aug. 6, 2015 to Sept. 4, 2026). Precision from an AI-assisted spot check.

Source: Author’s dictionaries applied to the Murillo corpus; AI-assisted spot checks.

How to read: each bar is one topic; its length is the number of broadcasts with at least one valid match. The gray note by each name is the precision: the share of a sample of matched sentences judged to be about the topic, and the sample size. Country topics use the site’s country colors; other topics are gray.

How this was calculated

  1. Data: /data/murillo/topics.json, built from the author’s topic dictionaries and the corpus in section 1.
  2. Paragraph text is accent-folded and, unless a pattern is listed as case-sensitive, lower-cased. A match is discarded if it overlaps an exclusion pattern, if its sentence matches a sentence exclusion (unless the sentence also matches an “unless” pattern), or if a date-limited rule removes it for that broadcast date (strip below). The hostile-US topic counts only sentences that match the US dictionary and a hostile cue.
  3. Bar length = broadcasts_matched: broadcasts with at least one valid match in their unmasked paragraphs, out of 2,404. The table also gives matches, the number of valid pattern matches (a sentence can hold more than one; for the hostile-US topic, one per sentence).
  4. Precision = correct ÷ sample size (precision.correct / precision.n). The sample: matched sentences (one per sentence), split by broadcast year, with the same number drawn at random from each year; each was read in context by an AI reader (Claude) on Sept. 26, 2026. Rounded to a whole percent.

Notes and caveats

Precision checks were AI-assisted.

The 2,404 broadcasts searched are the corpus from August 2015 on; the one 2012 transcript is left out.

Precision measures false matches only. It says nothing about mentions the patterns miss.

Where a sample exposed an error, the pattern was fixed after the sample; the note in the full table says so for each topic.

Sources

Author’s topic dictionaries applied to the Murillo broadcast corpus (section 1); precision spot checks of Sept. 26, 2026.

TopicGroupBroadcasts matchedValid matchesPrecisionSample correctSample size
RussiaPartners & rivals4591,691100%4848
China (PRC)Partners & rivals4261,19196%4345
TaiwanPartners & rivals381928100%4949
IranPartners & rivals102274100%4242
CubaPartners & rivals4231,08098%4748
VenezuelaPartners & rivals3971,23396%4648
United States (any mention)United States5681,22898%4748
Hostile references to the USUnited States7211994%4851
Sanctions vocabularyUnited States375798%5253
Daniel OrtegaLeadership & legitimacy1,8963,614100%4848
Laureano Ortega MurilloLeadership & legitimacy181225100%4545
Co-presidency vocabularyLeadership & legitimacy2026100%2626
"Pueblo Presidente"Leadership & legitimacy446740100%4141
"Padre Celestial"Leadership & legitimacy374491100%4848
"Traitors & coup-plotters"Enemies18041296%4345

Figure 2 above shows how many broadcasts each dictionary matched and the precision found in spot checks.

Rules that change on a known date

Some words changed meaning on a known date, so their exclusions apply only in a window. The bar shows when each rule is in force across the corpus.

  1. China (PRC)Before Dec. 10, 2021

    Before Nicaragua switched recognition to Beijing, phrases such as “Gobierno de China” or “República de China” meant Taiwan and are excluded; so is any China mention in a paragraph that also names Taiwan or its ambassador.

  2. TaiwanFrom Dec. 10, 2021

    After the switch, “República de China” without “Taiwán” is a slip for the People’s Republic, so it counts under China, not Taiwan.

  3. Co-presidency vocabularyBefore Jan. 1, 2025

    Before the 2025 reform the word referred only to co-chairs of commissions and forums, so no match before Jan. 1, 2025 counts.

Exclusions drop look-alikes, such as streets, schools or people that share a country’s or leader’s name. The table shows how many broadcasts each dictionary matched. A precision estimate (the share of matches a reader judged to be genuinely about the topic) is shown only where such a check was made; otherwise the counts are unchecked dictionary matches.

Full dictionary table: every search pattern, exclusion and date rule
Topic dictionaries used on the Murillo’s Broadcasts page
TopicWhat countsWhat is excludedBroadcasts matched (of 2,404)Precision estimate
Partners & rivals
Russia Any reference to contemporary Russia: the country, its people or state, Moscow, the Kremlin, or senior Russian officials (Putin, Lavrov, Shoigu, Patrushev and others) and Russian state brands (Sputnik, ROSATOM, Russia Today). References only to the Soviet Union or the Second World War are not counted.
Search patterns: \brusia\b \brus[oa]s?\b \bmoscu\b \bkremlin\b \bputin\b \blavrov\b \bsho[iy]gu\b \bpatrushev\b \bmedvedev\b \bvolodin\b \bmatvi[iy]enko\b \bbelou?sov\b
Show 6 more\bzajarova\b \bzakharova\b \bsputnik\b \brosatom\b \brussia today\b \bsan petersburgo\b
Sentences about the Second World War or the Soviet Union are dropped unless they also name the Russian Federation, its government, president, foreign minister, embassy or a delegation (for example, Victory Day greetings to Putin still count).
Show 1 exclusion patterngran guerra patria|union sovietica|\burss\b|sovietic|ejercito rojo|segunda guerra mundial|stalingrado|leningrado
459 100% (n = 48)
All sampled sentences referred to contemporary Russia.
China (PRC) References to the People's Republic of China: China, Chinese, Beijing, Xi Jinping, Wang Yi. Taiwan's official names ("China-Taiwán", "China (Taiwán)", and before the December 2021 switch "República de China") count under Taiwan, not here. Chinandega (a Nicaraguan department) never matches.
Search patterns: \bchinas?\b \bchin[oa]s?\b \bpekin\b \bbeijing\b \bxi\s*jin\s*g?\s*ping\b \bwang yi\b
Excludes every Taiwan form (China-Taiwán, China Taiwán, China (Taiwán), República de China), the nickname "El Chino" (e.g., the martyr Edmundo "El Chino" Pérez) and references to the Chinese language ("idioma chino", "chino-mandarín"), which before 2021 were Taiwan-funded courses. Before the December 2021 switch, Nicaragua recognized Taiwan, so pre-switch phrases such as "Pueblo y Gobierno de China" or "Embajada de China", and any China mention in a paragraph that also names Taiwan or its ambassador, are treated as Taiwan and excluded. Also excludes Traditional Chinese Medicine, the rambutan fruit ("mamón chino") and similar non-country uses.
Date-limited rule
Before Dec. 10, 2021also excluded: \brepublica de china\b \b(pueblo|gobierno|embajada|embajador|embajadora|cooperacion|mision|mision tecnica|republica) (y (el )?gobierno |y (el )?pueblo )?(hermano )?de china\b(?!\s+popular)
paragraphs matching taiwan|taipei|\btsai\b|jaime wu|embajador jaime|chin mu wu are not counted for this topic
Show 12 exclusion patterns\bchin[ao][\s-]*\(?\s*taiwan \bchima[\s-]*taiwan \brepublica de china[\s-]*\(?\s*taiwan \bchina nacionalista\b taiwan[\s,)-]*\(?\s*china\b \bel chino\b \bchino[\s-]*mandarin \bidioma chino\b \bmandarin[\s-]*chino\b \bmamon(es)? chinos?\b \bmedicina (tradicional )?china\b \bcultura popular china\b
426 96% (n = 45)
Errors: an ancient Chinese poet (culture) and Chinese fireworks (a product). Pre-2021 hits that meant Taiwan were removed by the date rule before this sample.
Taiwan References to Taiwan (the Republic of China): Taiwán, taiwanés/taiwanesa, Taipei, President Tsai Ing-wen, "China-Taiwán", and "República de China" before the December 2021 switch.
Search patterns: \btaiwan\b \btaiwanes(a|es|as)?\b \btaipei\b \btsai\b \brepublica de china\b \bformosa\b
Excludes a guava variety ("guayaba taiwanesa"). From Dec. 10, 2021, "República de China" without "Taiwán" is a slip for the People's Republic (e.g., "Presidente de la República de China, Hermano Xi Jinping") and counts under China, not Taiwan. After the switch most remaining hits are Beijing's "one China" line (e.g., Dec. 3, 2025).
Date-limited rule
From Dec. 10, 2021also excluded: \brepublica de china\b(?![\s-]*\(?\s*taiwan)
Show 1 exclusion pattern\bguayaba taiwanesa\b
381 100% (n = 49)
Two samples after moving post-2021 "República de China" slips to China.
Iran References to Iran: Irán (capitalized, to avoid the Spanish verb "irán", "they will go"), iraní/iraníes, Tehran, and named Iranian leaders (Khamenei, Khomeini, Raisi, Rouhani, Ahmadinejad, Pezeshkian, Zarif, Araghchi, Soleimani).
Search patterns: \birani(es)?\b \bteheran\b \b(jamenei|khamenei|jamenei|jomeini|jomeiny|khomeini)\b \braisi\b \b(rohani|rouhani)\b \bahmadine[jy]ad\b \bpezeshkian\b \b(araghchi|araqchi)\b \b(soleimani|suleimani)\b \bzarif\b \bIran\b \bIRAN\b
Only capitalized "Irán" counts, because lower-case "irán" is almost always the verb. Excludes the Managua clinic "Policlínico Iraní".
Show 1 exclusion pattern\bpoliclinico irani\b
102 100% (n = 42)
Stratified random sample of matched sentences across years, each read in context by an AI reader (Claude), Sept 26, 2026.
Cuba References to Cuba: Cuba, cubano/a(s), Havana, Díaz-Canel, and Fidel or Raúl Castro when named as such ("Fidel Castro", "Comandante Fidel").
Search patterns: \bcuba\b \bcuban[oa]s?\b \bhabana\b \bdiaz[\s-]*canel\b \bfidel castro\b \braul castro\b \bcomandante (en jefe )?fidel\b \bfidel y raul\b
Excludes Nicaraguan places, schools and centers named after Cuba or Fidel (Barrio Cuba, Centro Educativo República de Cuba, Centro de Desarrollo Tecnológico "Comandante Fidel" and similar). "Fidel" alone is not counted, because other Fidels (e.g., Nicaraguan officials) appear often.
Show 4 exclusion patterns\b(parque|tiangue|complejo( deportivo)?|estadio|centro( de desarrollo tecnologico| educativo| escolar| tecnologico)?|colegio|escuela|instituto|barrio|reparto|rotonda|avenida|plaza|mercado|cancha|biblioteca|residencial|colonia|comunidad|comarca|villa|calle|asentamiento|puente|cooperativa)\s*["“”]?\s*(republica de )?(cuba|(comandante )?fidel( castro)?)\b \brepublica de cuba,?\s+(en|de) (leon|managua|masaya|granada|chinandega|esteli|matagalpa|jinotega|rivas|carazo|boaco|chontales|madriz|nueva segovia|tipitapa|ciudad sandino)\b \bcuba libre\b \b(parque|centro|colegio|escuela|instituto|auditorio|cancha|biblioteca|fundacion|plaza|mercado)\b[^.;:“"]{0,45}[“"]\s*(comandante )?(fidel|cuba)
423 98% (n = 48)
The one error (a Nicaraguan center named "Comandante Fidel") was fixed after the sample.
Venezuela References to Venezuela: Venezuela, venezolano/a(s), Caracas, Nicolás Maduro, Cilia Flores, Delcy Rodríguez, Hugo Chávez (as "Hugo Chávez" or "Comandante Chávez").
Search patterns: \bvenezuela\b \bvenezolan[oa]s?\b \bcaracas\b \bnicolas maduro\b \b(presidente|hermano|companero) (nicolas )?maduro\b \bcilia flores\b \ba nicolas, a cilia\b \bnicolas y cilia\b \bdelcy\b \bhugo chavez\b \bcomandante (hugo )?chavez\b \bde bolivar, de chavez\b
Excludes the Managua avenue "Avenida de Bolívar a Chávez" and the many Nicaraguan parks, markets, stadiums, schools, neighborhoods and centers named after Hugo Chávez or Venezuela (e.g., Parque Hugo Chávez, Villa Venezuela, Centro de Desarrollo Tecnológico "Comandante Hugo Chávez"). "Maduro" counts only with a first name or title, because "maduro" is also an adjective ("ripe", "mature").
Show 3 exclusion patterns\bde bolivar a chavez\b \b(parque|tiangue|complejo( deportivo)?|estadio|centro( de desarrollo tecnologico| educativo| escolar| tecnologico)?|colegio|escuela|instituto|barrio|reparto|rotonda|avenida|plaza|mercado|cancha|biblioteca|residencial|colonia|comunidad|comarca|villa|calle|asentamiento|puente|cooperativa|fundacion|auditorio|sector)\s*["“”]?\s*(republica (bolivariana )?de )?(venezuela|(comandante )?(hugo )?chavez)\b \b(parque|tiangue|complejo|estadio|centro|colegio|escuela|instituto|auditorio|cancha|biblioteca|fundacion|plaza|mercado)\b[^.;:“"]{0,45}[“"]\s*(comandante )?(eterno,? )?(hugo )?chavez
397 96% (n = 48)
Errors: two Nicaraguan neighborhoods named after Hugo Chávez; one pattern was fixed after the sample.
United States
United States (any mention) Any mention of the United States: Estados Unidos, EE.UU., norteamericano/a(s), estadounidense(s), yanqui(s), gringo(s), Washington, the White House.
Search patterns: \bestados unidos\b \bee\.?\s?uu\.? \bnorteamerican[oa]s?\b \bestadounidenses?\b \byanquis?\b \bgringos?\b \bwashington\b \bcasa blanca\b
Excludes Mexico's formal name ("Estados Unidos Mexicanos") and currency ("dólares estadounidenses/norteamericanos").
Show 3 exclusion patterns\bestados unidos mexicanos\b \bdolares (estadounidenses|norteamericanos)\b \byanqui \d
568 98% (n = 48)
The one error (road sites "Yanqui 1, Yanqui 2" in Jinotega) was fixed after the sample.
Hostile references to the US A sentence that mentions the United States (as defined above) AND contains a hostile cue: imperio/imperialismo, agresión, injerencia, criminal, genocidio, bloqueo, sanciones, amenaza, intervención, saqueo. Sentences about history before the 1979 revolution (Walker and San Jacinto, the 1912-1933 occupation and Sandino's war, Hiroshima) are excluded, as are weather bulletins; grievances about the 1980s Contra war and the 1986 World Court ruling are kept because she presents them as live claims.
Counted only in passages that also match “United States (any mention).”
Search patterns: \bimperi(o|os|al|ales|alismo|alista|alistas)\b \bagresi\w* \binjeren\w* \bcriminal\w* \bgenocid\w* \bbloqueo\w* \bsancion\w* \bamenaz\w* \bintervenc\w* \bintervencionis\w* \bsaque[oa]\w*
This cue method finds explicit vilification. It misses policy critiques phrased without insults (for example, criticism of tariffs or of the treatment of migrants), so it undercounts criticism in 2025-26. See the separate estimate by administration below the explorer.
Show 2 exclusion patterns\bsandino\b|\bwalker\b|filibuster|\b18[0-9]{2}\b|\b19[0-4][0-9]\b|san jacinto|zeledon|guerra nacional|ejercito defensor|primeras? derrotas?|imperio (britanico|espanol)|hiroshima|nagasaki|bomba atomica|siglo xix|somoza|somocis huracan|tormenta tropical|onda tropical|depresion tropical|\bciclon|centro nacional de huracanes|\bnoaa\b|\bsismo|terremoto|\bclima\b|tornado|nevada|incendios? forestal
72 94% (n = 51)
Precision judged as "the sentence expresses hostility toward the US government or its policies (not pre-1979 history)". Errors were historical references; history cues were widened after the sample.
Sanctions vocabulary The word family sanción/sanciones/sancionar used for international sanctions, plus "medidas coercitivas unilaterales" (the regime's preferred term). Covers sanctions on Nicaragua, on allies, or in general. Domestic uses (penalties in regulations, enacting a law, criminal penalties) are excluded.
Search patterns: \bsancion\w* \bmedidas (unilaterales, )?coercitivas\b
Drops sentences where 'sancionar' means enacting a law, and sentences about domestic regulations, fines, traffic, fireworks, forest fires, sports, criminal penalties or the Genocide Convention's 'prevent and punish', unless they also mention unilateral, 'so-called' or US measures.
Show 3 exclusion patternsprevencion y (la )?sancion del (delito|crimen|genocidio) sancion\w* (la|esta|dicha|una|nueva|las|estas) (ley|leyes|reforma)|(ley|leyes|reforma|decreto)\w* \w*\s?sancionad|sancion\w* y (promulg|public)|promulg\w* y sancion \bmulta|\binfraccion|infractor|incumpl|transito|conductor|arbitr|\bfifa\b|\bpenal(es)?\b|\bpenas\b|codigo|delito|normativa|reglamento|\bresolucion|polvora|incendio|reincident|\bley \d+|prevenir y sancionar|probidad|administrativamente|sanciona la traicion
37 98% (n = 53)
All 53 matched sentences read. The one error (the Genocide Convention's "prevención y la sanción") was fixed after the check.
Leadership & legitimacy
Daniel Ortega Daniel Ortega named by his usual titles: "Comandante Daniel", "Daniel Ortega", "Presidente Daniel", "Co-Presidente Daniel", "Comandante-Presidente Daniel", "Compañero Daniel". Other Daniels, including his son Daniel Edmundo Ortega Murillo, are excluded.
Search patterns: \bcomandante daniel\b \bdaniel ortega\b \bpresidente daniel\b \bco-?\s?presidente daniel\b \bcomandante[\s-]+presidente daniel\b \bpresidente[\s-]+comandante daniel\b \bcompanero daniel\b
Excludes Daniel Edmundo Ortega Murillo and other Daniels (no bare 'Daniel').
Show 2 exclusion patterns\bdaniel ortega murillo\b \bdaniel edmundo\b
1,896 100% (n = 48)
Stratified random sample of matched sentences across years, each read in context by an AI reader (Claude), Sept 26, 2026.
Laureano Ortega Murillo References to Laureano Ortega Murillo, the co-presidents' son, presidential adviser for investment, trade and international cooperation (from March 2023 also called the President's special representative for relations with Russia).
Search patterns: \blaureano\b
Excludes the Managua neighborhood and hero "Laureano Mairena" and an unrelated "Laureano Mateo".
Show 1 exclusion pattern\blaureano (mairena|mateo)\b
181 100% (n = 45)
Read after excluding the Managua neighborhood "Laureano Mairena", which was the main source of error in a first sample.
Co-presidency vocabulary "Co-Presidente", "Co-Presidenta", "Co-Presidencia" (with or without hyphen), the titles created by the 2025 constitutional reform. Co-chairs of commissions or forums are excluded.
Search patterns: \bco-?\s?presiden(te|ta|tes|tas|cia)\b
Counts only from Jan. 1, 2025: before the 2025 reform the word referred only to co-chairs of commissions and forums (e.g., the intergovernmental commission with Russia). Later co-chair uses are also dropped.
Date-limited rule
Before Jan. 1, 2025also excluded: \bco-?\s?presiden\w*
Show 1 exclusion patternco-?\s?presiden\w* (de|del) (la|el) (comision|foro|comite|grupo|consejo|mesa|conferencia|cumbre|reunion)|comision (mixta|intergubernamental)
20 100% (n = 26)
Every matched sentence read (only 26 exist after the 2025 date rule).
"Pueblo Presidente" The slogan "Pueblo Presidente" ("the People as President"), written with a space or hyphen.
Search patterns: \bpueblo[\s-]*presidente\b
None 446 100% (n = 41)
Stratified random sample of matched sentences across years, each read in context by an AI reader (Claude), Sept 26, 2026.
"Padre Celestial" The devotional phrase "Padre Celestial" ("Heavenly Father").
Search patterns: \bpadre celestial\b
None 374 100% (n = 48)
Includes Bible verses she reads aloud; the phrase itself is always the devotional one.
Enemies
"Traitors & coup-plotters" The regime's labels for opponents: golpista/golpismo, vendepatria(s), traidor(es), "traición a la patria", apátrida(s). These words are also used occasionally about foreign events (e.g., coups elsewhere).
Search patterns: \bgolpis\w* \bvende-?\s?patrias?\b \btraidor(a|es|as)?\b \btraicion a la patria\b \bapatridas?\b
Drops sentences about Sandino's era and other history (e.g., Sandino calling 1920s politicians 'vendepatrias') and Judas. Quoted poems that use these words about contemporary opponents still count.
Show 1 exclusion pattern\bsandino\b|adolfo diaz|chema moncada|\bsacasa\b|\bwalker\b|filibuster|\bjudas\b
180 96% (n = 45)
Errors: a quotation of Sandino without his name in the sentence, and "traitors to Christ".

15 topics. Patterns are matched in Spanish, without regard to accents or (unless noted) capitalization; \b marks a word boundary. A date-limited rule applies only to broadcasts inside its window (“from” dates inclusive, “before” dates exclusive).

Technical note on matching

Topic dictionaries for the Murillo broadcast explorer. All patterns are Python regular expressions applied to paragraph text that has been accent-folded (á->a, ñ->n) and, unless listed under case_patterns, lower-cased. A match counts only if it does not overlap an 'exclude' match and its sentence is not removed by 'exclude_sentence' (unless the sentence also matches 'unless_sentence'). A broadcast counts toward a topic if at least one valid match occurs in its unmasked paragraphs.

03

Topic-mix validation

The “what does she talk about” measure assigns every paragraph to one of eleven broad subjects (public works, faith, foreign affairs and so on) using a keyword classifier, then reports each subject’s share of her words per year. A blind second reading agreed with it on 68.5% of paragraphs.

Figure 3

Of ten shifts in her topic mix shown here, five are fully robust and two are not.

Change in a subject’s share of her words, in percentage points: raw, corrected for the classifier’s measured error, and the 95% bootstrap interval of the corrected change.

  • Corrected change
  • Raw change
  • 95% interval
  • Robust
  • Direction holds; size uncertain
  • Not robust

Source: Author’s classifier and blind validation, Murillo corpus.

How to read: each row is one shift. The filled dot is the change after correcting for the classifier’s measured error, the hollow dot is the raw change, and the bar is the 95% interval. A bar that crosses the zero line cannot rule out no change. Color gives the verdict.

How this was calculated

  1. Data: /data/home/validation.json. The shift estimates and verdicts are transcribed from the author’s topic-mix validation report (its section 3), which tested more shifts than the ten shown here.
  2. Raw change: the subject’s share of her words in the later period minus its share in the earlier period, in percentage points, as scored by the keyword classifier (each paragraph assigned to its top-scoring subject).
  3. Corrected change: each period’s shares are corrected with the confusion matrix from the 200-paragraph blind check (corrected share = sum over classifier labels of raw share × P(blind reader’s subject | classifier label), weighted by words), and the change is recomputed.
  4. Interval: the 95% bootstrap interval of the corrected change, from 2,000 resamples of the 200-paragraph validation sample.
  5. Verdict “Robust”: (a) keeps its sign and a size of at least 2 points under five scoring variants, (b) survives the error correction, (c) the interval excludes zero, and (d) it is more than twice the largest difference between two outlets’ copies of the same broadcast. “Direction only”: (b) and (c) hold but the size is uncertain. In the chart, “Direction only” and the two rows marked “size uncertain” share one color (“Direction holds; size uncertain”); the exact verdict is in the tooltip and the table.

Notes and caveats

The blind second reading was an independent AI-assisted reading, not a human panel.

Treat the topic shares as evidence of direction and rough size, not as exact percentages.

Sources

Author’s keyword classifier and blind validation of 200 paragraphs from the Murillo broadcast corpus.

ShiftRaw change (pts)Corrected (pts)95% interval, low95% interval, highVerdict
News desk (weather, disasters, health bulletins), 2017 to 2026−12.0−9.5−11.8−6.4Robust
Faith, 2017 to 2018-19+11.8+8.2+5.3+10.4Robust
Faith, 2022-25 to 2026+6.7+4.7+3.1+6.0Robust
Services and works, 2017 to 2022-25+6.7+4.2+1.7+6.2Robust
Revolution, party and history, 2017 to 2026+7.3+4.1+0.8+7.4Direction robust; size uncertain
Foreign affairs, 2024 to 2026−7.0−4.1−6.3−1.9Robust; size uncertain
Economy, 2017 to 2026−4.5−3.3−4.3−1.5Robust
Foreign affairs, 2017 to 2026−4.8−1.4−4.0+2.0Not robust
Culture and festivities, 2017 to 2026+2.3+1.0−0.6+2.3Not robust
Security and "enemies", 2024 to 2026−2.2−1.2−2.0−0.3Direction only
  • 200paragraphs read blind (20 per year)
  • 68.5%agreement; Cohen’s kappa 0.62
  • 84%matched the reader’s first or second choice

Figure 3 above shows which measured shifts in her topic mix are robust.

Known weaknesses

  • The classifier under-detects weather and disaster content in 2017–18, so the decline of the “news desk” is, if anything, larger than shown.
  • “Ritual greetings” absorbs slogan-heavy paragraphs; its level is overstated by several points and shows no reliable trend.
  • Foreign affairs is under-detected (recall about 40%), so its levels are low; the fall from 2024 to 2026 is robust, but comparisons with 2017 are not.
  • 2017–2021 transcripts come from Canal 4 and 2024–2026 from El 19 Digital. Copies of the same broadcast from two outlets differ by at most about one point per subject.
Learn more: the blind check and what “robust” means

To test it, a second coder read 200 fresh paragraphs (20 per year) without seeing the classifier’s labels. The second coder was an independent AI-assisted reading, not a human panel. The two agreed on 68.5% of paragraphs (Cohen’s kappa 0.62, conventionally “substantial” agreement). In 84% of paragraphs the classifier’s label matched either the blind reader’s first or second choice, so most disagreements involve paragraphs that mix two subjects. An earlier, non-blind check by the classifier’s builder had reported 83%; the site uses the lower, blind figure.

What that means for the charts. Treat the topic shares as evidence of direction and rough size, not as exact percentages. A shift is called robust only if it (a) keeps its sign and a size of at least 2 points under five scoring variants, (b) survives correction for the classifier’s measured error, (c) has a 95% bootstrap interval that excludes zero, and (d) is more than twice the largest difference between two outlets’ copies of the same broadcast. “Direction only” means (b) and (c) hold but the size is uncertain.

Learn more: how well each subject was detected

Precision: of the paragraphs the classifier put in a subject, the share the blind reader also put there. Recall: of the paragraphs the blind reader put in a subject, the share the classifier found. Small counts make several of these rough.

SubjectBlind reader (n)Classifier (n)PrecisionRecall
Government services and public works666685%85%
Weather, disasters and health bulletins251782%56%
Ritual greetings and exhortation243459%83%
Foreign affairs231464%39%
Culture, festivities and sports171362%47%
Faith142060%86%
Security and "enemies"121070%58%
Economy6862%83%
Revolution, party and history6757%67%
State, law and elections4367%50%
Unclassified380%0%
04

Lending data

The NICA Act section compares lending to Nicaragua’s government before and after the Nicaragua Human Rights and Anticorruption Act of 2018 (NICA Act, P.L. 115-335). All amounts are nominal US dollars. Approvals (commitments at board approval) and disbursements (money actually paid out) are shown separately and never added together; disbursements lag approvals by years.

DatasetWhat it measuresCoverage
World Bank Projects APIWorld Bank (IDA/IBRD) project approvals for Nicaragua, by board-approval year2010–Sept. 2026
IDB Projects DatasetInter-American Development Bank (IDB) sovereign-guaranteed loan approvals; technical-cooperation grants kept separateDownloaded Sept. 25, 2026
CABEI open data (“Préstamos”)Central American Bank for Economic Integration (CABEI) loan approvals and disbursements, public and private sectorFiles dated Sept. 14, 2026
World Bank International Debt Statistics (IDS)Disbursements, commitments and net flows on public and publicly guaranteed debt, by creditor (IDA, IDB, CABEI, OPEC Fund, IMF, China)Through 2024 (released Dec. 2025); 2025 projections dropped
US Treasury MDB voting recordsHow the US executive director voted on each Nicaragua operation at the multilateral development banks (MDBs)2016–Aug. 2025
Treasury quarterly report on IMF lending (Q4 2020)US position on the International Monetary Fund’s (IMF) 2020 emergency financingOct.–Dec. 2020
Nicaragua Ministry of Finance (MHCP) annual debt-policy evaluationsNew external loan contracts by creditor, including Chinese supplier credits2021–Q1 2026

Why NICA’s effect cannot be isolated

  • Timing. IDB and World Bank sovereign lending stopped in 2018, before the Act was signed on December 20, 2018. Loans that were never proposed leave no vote.
  • Other rules. Other US rules (trafficking-in-persons restrictions, human-rights voting law, the Pelosi amendment on environmental disclosure and Executive Order 13851) pushed in the same direction.
  • What was approved. The only approvals after 2018 were COVID-19, hurricane and disaster operations.
  • Substitution. CABEI and Chinese contractors replaced much of the lost volume.
  • An upper bound. The “about $1.6 billion” foregone-approvals figure is therefore an upper bound computed against the 2010–2017 pace, not an estimate of NICA’s effect.
  • Not public. Treasury’s reports and briefings required by the Act are not public.
05

Sanctions lists

Sanctions status was checked against the full published list files, not press summaries. “Not listed” means not found on any of these public lists as of the dates below.

ListPublisherAs of
Specially Designated Nationals (SDN) and Consolidated listsUS Treasury, Office of Foreign Assets Control (OFAC)Published Sept. 23, 2026
EU consolidated financial sanctions list (Nicaragua regime, Reg. 2019/1716)European CommissionGenerated Sept. 22, 2026
UK Sanctions ListUK Foreign, Commonwealth & Development OfficeReport date Sept. 21, 2026
Consolidated Canadian autonomous sanctions listGlobal Affairs CanadaLatest listing Sept. 22, 2026
Section 7031(c) public visa designations (a provision of the annual State Department appropriations act)US Department of State press releasesSearched through Sept. 25, 2026

What the lists cannot show

The State Department can bar officials from entering the United States privately under Section 7031(c) and Presidential Proclamation 10309; State said in June 2026 that visa restrictions cover over 2,350 Nicaraguan officials and family members. Those names are not public, so an official “not on a public list” may still be barred from entry.

Learn more: how names were matched

Matching. Names were normalized (accents removed) and matched on a surname plus a second surname or given name, across primary names and all aliases, on every sanctions program, not only Nicaragua programs. A looser surname-only pass was run over entries mentioning Nicaragua, and distinctive tokens were searched in the raw files. False positives (unrelated people with similar names) were removed by hand. OFAC designation dates come from the list’s own event records.

06

Gold trade data

Four sources: importer-reported trade from UN Comtrade and the US Census Bureau, Nicaragua’s own export totals from its central bank, and the Federal Register for US import-prohibition authority.

  • UN Comtrade A

    US-reported imports from Nicaragua of HS 7108 (unwrought gold, almost entirely bullion and doré), annual and monthly, in dollars. Canadian and Swiss imports from the same source show where the rest of Nicaragua’s gold goes.

  • US Census Bureau A

    Its International Trade API, for the state each import is declared for, the port where it clears customs, and 10-digit tariff lines, whose quantities give the gold content in kilograms. The 2025 Utah figure matches the Kem C. Gardner Policy Institute’s published Utah International Trade, 2025.

  • Central Bank of Nicaragua (BCN)

    Export reports, for Nicaragua’s own export totals. The 2025 gold export total ($1.97 billion) is as reported by La Prensa citing the BCN C, because the BCN’s own tables could not be retrieved. Nicaragua’s coding of export destinations differs from what importing countries record, so destination shares use importer-reported data.

  • Federal Register and OFAC notices

    For the import-prohibition authority added by Executive Order 14088 (Oct. 24, 2022). No determination under it has been published as of Sept. 26, 2026.

Trade values are nominal. US imports are reported at CIF value (including freight and insurance) unless labeled customs value; the difference for Nicaraguan gold is under 0.1%.

07

Records of Chinese, Russian and Iranian activity

The Russia, China & Iran section draws on a catalogue of 130 public-record items (66 Chinese, 32 Russian, 32 Iranian) with events from January 1, 2018 to September 24, 2026 (a few earlier assets appear only as context), compiled from provider and recipient government documents, company filings, legal texts, Nicaraguan budget reports, and independent and regime media. Taiwan is excluded from Chinese totals.

Records in the catalogue, by provider (130 in all). A record is not a delivered asset; see the first rule below.

  • Counts are records, not delivered assets. A record can be a project, program, agreement, delivery series, company portfolio, proposal or a lead that was checked and excluded. Trend charts count only eligible records with a dated event in a given year, and recent years are better documented than earlier ones.
  • Money measures are never summed across types. Loan commitments, contract values, disbursements, budget accruals and debt service are different things. Chinese and Russian loan figures come from the Finance Ministry’s registered contracts.
  • An official record shows what it records. A signed agreement or announced handover is not proof of delivery or use. Equipment origin or dual-use potential does not establish foreign access, intelligence collection or basing; those need their own evidence.
  • The catalogue is not exhaustive. Searches expanded the original leads substantially, but undisclosed transactions, complete arms-transfer registers and customs records were not available.
08

Security officials profiles

The Military & Police section draws on the author’s compiled profiles of Army, National Police, Interior Ministry and related posts compiled in September 2026 from open sources.

  • Presidential decrees (acuerdos presidenciales) and La Gaceta
  • Official Army and Police notes
  • OFAC and State Department releases
  • UN reports, including the Group of Human Rights Experts on Nicaragua (GHREN)
  • Named investigative outlets: Confidencial, Divergentes, La Prensa, Artículo 66, EFE
  • Each post-holder, date and relationship carries its own evidence tier. Dates of lower precision (month or year only) are flagged, and tenure bars use them as approximations.
  • Sanctions entries in those profiles were corrected against the lists described in section 5; for example, one designation date was moved from Dec. 12 to Nov. 7, 2019, and two officials previously shown as EU-listed are not.
  • Projected term ends under the September 2026 constitutional reform are the author’s calculations from the reform’s transitional clause, which counts seven years from each chief’s latest appointment. They assume ratification in January 2027 and are shown as ranges where the appointment date is ambiguous.
  • Public officials and sanctioned persons are named. Relatives who are private individuals are described, not named, and no personal identifiers beyond those in official sanctions listings are published.
09

Political prisoners and purges

The Known Instances of Political Imprisonment page and the purge section of the Military & Police page combine the author’s compiled records on political prisoners (research cutoff September 26, 2026) with a second, AI-assisted research stage conducted on September 26, 2026.

  • Monitor registers: the Mechanism for the Recognition of Political Prisoners and Asociación Memoria y Justicia
  • UN Group of Human Rights Experts (GHREN), UN human rights office (OHCHR), IACHR
  • US Treasury, State Department, USTR and Federal Register; EU, UK, Canadian and Swiss lists; OAS and Holy See records
  • The Murillo broadcast corpus (section 1)
  • Named outlets: Confidencial, La Prensa, Divergentes, Artículo 66, 100% Noticias, EFE, AP
  • Stocks are not flows. Counts from different organizations are never added or subtracted; overlapping totals are never summed; no death rate, regression or risk score is computed.
  • Independent review. Ten studies (eight on prisoners, two on purges) were each checked against primary sources by a separate reviewer; the published findings apply every correction and omit claims that failed. This stage has not been verified by a person.
  • Privacy. Only public figures are named: officials, officers, party and Indigenous leaders, bishops and prominent opposition figures. Everyone else is described by role, and the data build fails if a withheld name appears in a published file.
10

Evidence tiers

Claims across the site carry a letter grade for the strength of the underlying evidence.

A

Official or primary

An official dataset, legal text, government record or primary document read directly: for example a sanctions list file, a lender’s dataset, La Gaceta, or a transcript’s exact wording.

B

Reputable secondary

Reporting by named, established outlets or research institutions, ideally more than one. Attributed in the text (“according to Confidencial”).

C

Single report or regime claim

One press report, or the Nicaraguan government’s own claim about itself (for example trainee numbers or handovers). Always attributed; never stated as fact.

D

Calculation or inference

The author’s computation or analytic judgment built on A–C inputs, such as projected term ends or a foregone-lending gap. The inputs and method are stated.

Counts from Murillo’s broadcasts are a separate category: the transcript text itself is primary, but the counts and classifications are produced by dictionary search plus AI-assisted reading and checked by the author; minor errors may remain.

11

Corrections and how to cite

Errors are corrected promptly. A correction note with its date is added beneath the affected figure or paragraph, and the change is not made silently.

How to cite the site

Guerra, Gil. Data for US-Nicaragua Policy. 2026. nicaragua-research.com/

How to cite a single figure

Guerra, Gil. “[Figure title].” Data for US-Nicaragua Policy. 2026. [Figure link]

Add the figure number and title, and the figure’s own link (use the Link button under the figure).

Corrections. Numbers that were revised during the research are shown in their corrected form only; for example, the US share of Nicaragua’s gold exports is roughly a third by importer records, and topic-classifier agreement is 68.5%, not the earlier 83%. To report an error, contact the author through Hudson Institute.

12

Data downloads

Every chart on the site has four actions beneath it. Figures and data may be reused with attribution.

  • TableShow data table: the exact values plotted, readable by screen readers.
  • CSVDownload data (CSV): the same table as a spreadsheet file.
  • ImageSave image (PNG): the chart with its title, legend and source line, ready for a slide or memo.
  • LinkCopy link to figure: a direct link that opens the page at that figure.

The processed data behind each page are also published as JSON files under /data/ on this site. Source files that are not public (for example working spreadsheets) are not republished.