Pulse backtest reportBeta

Summary

Verdict thresholds

For each expected (dimension, direction) row, the case passes if the absolute peak |Δ| within the ±90 day verdict window (relative to the case's event date) reaches the magnitude threshold in the right direction. Magnitudes:

• Moderate: |Δ| ≥ 1.0
• Severe: |Δ| ≥ 3.0
• Catastrophic: |Δ| ≥ 5.0

On August 15, 2021, the Taliban entered Kabul as the Islamic Republic of Afghanistan collapsed. President Ashraf Ghani fled the country; the elected government dissolved; the constitutional order ended. Within weeks the Taliban declared an Islamic Emirate, replaced the secular legal system with Sharia rule, dissolved the elected parliament, banned all political parties, and progressively excluded women from secondary education and most paid employment. The case is the spec's archetypal catastrophic-across-all-dimensions example.

Expected vs. computed

Trajectories

Sampled every 30 days from −180 to +365. Vertical accent line marks the case event date; the lighter band is the ±90-day verdict window. Dashed grid lines mark ±1 (the moderate threshold).

Notes

• Democratic Quality: Peak -8.07 on day 30 matched expected catastrophic negative.
• Rule of Law: Peak -5.93 on day 30 matched expected catastrophic negative.
• Rights & Freedoms: Peak -14.17 on day 60 matched expected catastrophic negative.

Last ran Apr 30, 2026

On January 8, 2023, supporters of former president Jair Bolsonaro stormed the National Congress building, the Supreme Federal Tribunal, and the Planalto Palace in Brasília, attempting to overturn the result of the October 2022 election that had brought Lula da Silva to power. The institutional response was swift: the military did not back the rioters, federal police and security forces secured the buildings within hours, more than 1,500 arrests followed, and the Supreme Court opened investigations into the organisers and into Bolsonaro himself. The case is a moderate-magnitude stability and rights stress test in which institutions held.

Expected vs. computed

Trajectories

Sampled every 30 days from −180 to +365. Vertical accent line marks the case event date; the lighter band is the ±90-day verdict window. Dashed grid lines mark ±1 (the moderate threshold).

Notes

• Stability: Peak -4.33 on day 0 matched expected moderate negative.
• Rights & Freedoms: Peak -1.94 on day 30 matched expected moderate negative.

Last ran Apr 30, 2026

After more than four years of negotiations in Havana, the Colombian government and the Revolutionary Armed Forces of Colombia (FARC-EP) signed a final peace agreement on November 24, 2016. The accord ended the longest-running armed conflict in the Western Hemisphere. The original agreement was rejected in an October 2 plebiscite by a narrow margin; a revised text was signed and ratified by Congress on November 30. This is a positive-direction case used to verify the pipeline correctly registers stabilisation.

Expected vs. computed

Trajectories

Sampled every 30 days from −180 to +365. Vertical accent line marks the case event date; the lighter band is the ±90-day verdict window. Dashed grid lines mark ±1 (the moderate threshold).

Notes

• Stability: Peak 6.18 on day 30 matched expected moderate positive.

Last ran Apr 30, 2026

On November 4, 2020, Prime Minister Abiy Ahmed launched a military operation in Ethiopia's Tigray region, beginning a civil war that lasted approximately two years and killed an estimated 600,000 people through combat, famine, and disease. The conflict involved federal forces, Eritrean troops, and Tigray Defense Forces; widespread atrocities including massacres of civilians, sexual violence as a weapon of war, and aid blockades that produced famine conditions. The November 2022 Pretoria peace agreement formally ended major combat operations.

Expected vs. computed

Trajectories

Sampled every 30 days from −180 to +365. Vertical accent line marks the case event date; the lighter band is the ±90-day verdict window. Dashed grid lines mark ±1 (the moderate threshold).

Notes

• Stability: Peak -6.50 on day 0 matched expected catastrophic negative.
• Rights & Freedoms: Peak -6.35 on day 30 matched expected catastrophic negative.

Last ran Apr 30, 2026

Hungary's democratic backsliding under successive Fidesz governments led by Viktor Orbán is the spec's archetypal gradual-erosion case. Unlike a coup, the trajectory plays out across more than a decade through legislative supermajorities, constitutional rewrites, judicial restructuring, media-ownership concentration, NGO targeting, and electoral-rule changes. The Pulse's single-event-window framework (−180 to +365 days from one anchor date) cannot capture the full arc; for backtesting purposes the case is anchored to December 30, 2011 — the date the cardinal-law judicial restructuring took effect, sitting between the April 2011 constitutional adoption (~day −250) and the January 2012 electoral law (~day +24). Readers should consult the Civica Index trend chart, which is the appropriate tool for slow-erosion cases.

Expected vs. computed

Trajectories

Sampled every 30 days from −180 to +365. Vertical accent line marks the case event date; the lighter band is the ±90-day verdict window. Dashed grid lines mark ±1 (the moderate threshold).

Notes

• Democratic Quality: Peak -5.71 on day 30 matched expected severe negative.
• Rule of Law: Peak -7.76 on day 0 matched expected severe negative.

Last ran Apr 30, 2026

Myanmar military (Tatmadaw) seized power on February 1, 2021, detaining elected NLD officials including Aung San Suu Kyi and declaring a year-long state of emergency. The coup ended a decade of partial democratic transition. Cascade events through the rest of 2021 — parliament dissolution, election annulment, party dismantling, show trials of the elected leadership, expansion of military tribunals to civilians — drove dimensional impact across democratic_quality, rule_of_law, and freedom_rights in addition to the headline stability rupture.

Expected vs. computed

Trajectories

Sampled every 30 days from −180 to +365. Vertical accent line marks the case event date; the lighter band is the ±90-day verdict window. Dashed grid lines mark ±1 (the moderate threshold).

Notes

• Democratic Quality: Peak -11.20 on day 90 matched expected catastrophic negative.
• Rule of Law: Peak -7.49 on day 60 matched expected catastrophic negative.
• Rights & Freedoms: Peak -8.32 on day 30 matched expected severe negative.

Last ran Apr 30, 2026

Members of Niger's presidential guard detained President Mohamed Bazoum on July 26, 2023, declaring a coup d'état and installing General Abdourahamane Tchiani as head of a new junta (Conseil National de Sauvegarde de la Patrie). Cascade events through late 2023 — constitution suspended, parliament dissolved, political party activities banned, treason charges filed against the deposed president, three-year transition timeline announced — drove dimensional impact across democratic_quality and rule_of_law in addition to the headline stability rupture.

Expected vs. computed

Trajectories

Sampled every 30 days from −180 to +365. Vertical accent line marks the case event date; the lighter band is the ±90-day verdict window. Dashed grid lines mark ±1 (the moderate threshold).

Notes

• Democratic Quality: Peak -8.18 on day 60 matched expected catastrophic negative.
• Rule of Law: Peak -4.14 on day 90 matched expected severe negative.

Last ran Apr 30, 2026

On October 15, 2023, a coalition of opposition parties led by Donald Tusk's Civic Coalition won enough parliamentary seats to end eight years of governance by the Law and Justice (PiS) party. The new coalition government, sworn in December 13, immediately began reversing PiS-era restrictions on judicial independence, public media, and civil society. The case is the spec's archetypal positive-direction example — a peaceful democratic transfer of power following a period of backsliding, and a stress test of whether the Pulse correctly registers improvements as well as deteriorations.

Expected vs. computed

Trajectories

Sampled every 30 days from −180 to +365. Vertical accent line marks the case event date; the lighter band is the ±90-day verdict window. Dashed grid lines mark ±1 (the moderate threshold).

Notes

• Democratic Quality: Peak 3.20 on day 60 matched expected moderate positive.

Last ran Apr 30, 2026

Sri Lanka's 2022 economic collapse — sovereign default, severe shortages of fuel, food, and medicine — triggered the largest peaceful protest movement in the country's modern history. Mass demonstrations across spring and summer 2022 culminated on July 9 when protesters stormed the presidential palace. President Gotabaya Rajapaksa fled the country and resigned days later. The case is a moderate-magnitude stability and rights stress test in which constitutional process largely held: Wickremesinghe was elected president by parliament, no military takeover occurred, civil-society institutions endured.

Expected vs. computed

Trajectories

Sampled every 30 days from −180 to +365. Vertical accent line marks the case event date; the lighter band is the ±90-day verdict window. Dashed grid lines mark ±1 (the moderate threshold).

Notes

• Stability: Peak -2.76 on day 0 matched expected moderate negative.
• Rights & Freedoms: Peak -6.53 on day 30 matched expected moderate negative.

Last ran Apr 30, 2026

On July 25, 2021, Tunisian President Kais Saied invoked Article 80 of the constitution to dismiss the prime minister, freeze parliament, and assume executive powers. Critics labeled the move a self-coup. Subsequent decrees further consolidated presidential power, including a 2022 referendum that adopted a new constitution removing parliamentary checks on the executive.

Expected vs. computed

Trajectories

Sampled every 30 days from −180 to +365. Vertical accent line marks the case event date; the lighter band is the ±90-day verdict window. Dashed grid lines mark ±1 (the moderate threshold).

Notes

• Democratic Quality: Peak -4.33 on day 30 matched expected severe negative.
• Rule of Law: Peak -4.33 on day 0 matched expected severe negative.

Last ran Apr 30, 2026

Cite this page

Civica. (2026). Civica Atlas Methodology — Pulse backtest report (Beta): Pulse backtest report. Civica Atlas. Retrieved May 7, 2026, from https://civicaatlas.org/civica-index/methodology/pulse/backtest