How The Abolition Of The Death Penalty Spread Nation By Nation

Amnesty International's annual report — the most widely cited source tracking the global picture — divides countries into four categories:

- Abolitionist for all crimes: 112 countries as of the most recent count. - Abolitionist for ordinary crimes only (retain it for military or exceptional offenses): 9 countries. - Abolitionist in practice (retain the law but haven't executed anyone in at least 10 years and have an established policy or practice of non-execution): 23 countries. - Retentionist: approximately 55 countries.

Total abolitionist in law or practice: 144. That's where the widely-cited number comes from.

Now, that second category matters. Hiding inside "retentionist" are countries that technically allow capital punishment but don't actually use it, or use it rarely. And inside "abolitionist in practice" are countries one political shift away from resuming executions. The boundary is porous. But the directional trend — measured by executions per year, number of abolitionist countries, and number of states ratifying anti-death-penalty treaties — has moved in one direction for seventy years.

The standard periodization, drawn from Roger Hood and Carolyn Hoyle's The Death Penalty: A Worldwide Perspective (now in its fifth edition, the definitive text on this), looks roughly like:

Wave 1 (1846–1945): Scattered national abolitions. Michigan abolished for non-treason offenses in 1846, the first jurisdiction in the English-speaking world to do so. Venezuela abolished for all crimes in 1863. Portugal in 1867. San Marino in 1865. Most were small states or peripheral reformist moves. The global practice remained universal.

Wave 2 (1945–1977): Post-war European abolition. West Germany (1949), Austria (1950, restored briefly, fully abolished 1968), the Vatican (1969), Sweden (1972, though already abolished in practice), Portugal (fully, 1976), Spain (1978). A direct response, scholars argue, to the industrial-scale state killing of the Nazi and fascist regimes. Europe recoiled.

Wave 3 (1977–2000): The democratization wave. Latin America (most of the continent abolishing as juntas fell), Central and Eastern Europe (post-1989 as a condition of EU accession), South Africa (1995 by Constitutional Court ruling in S v Makwanyane, one of the most cited judicial opinions on the death penalty globally). The Second Optional Protocol to the ICCPR (1989) gives the movement a multilateral legal scaffold.

Wave 4 (2000–present): African abolitionism, Asian gradualism, and moratorium politics. Rwanda (2007), Togo (2009), Gabon (2010), Benin (2012), Madagascar (2015), Chad (2020), Sierra Leone (2021), Central African Republic (2022), Zambia (2022), Ghana (2023). In Asia, Mongolia abolished in 2017. Malaysia abolished the mandatory death penalty in 2023. The UN General Assembly has passed biennial moratorium resolutions since 2007 with steadily increasing support — from 104 yes votes in 2007 to 125 in the most recent cycle.

This is not a straight line. There are reversals. The Philippines abolished, restored, abolished again. The US reinstated in 1976 after a four-year moratorium following Furman v. Georgia. Japan continues to execute. But the trend across decades is unmistakable.

This is where it gets interesting. The received wisdom — that countries abolish when their publics want them to, that abolition follows drops in violent crime, that scientific consensus on deterrence shifts opinion — is largely wrong. Or at least, seriously incomplete.

The scholars to read here are Carol Steiker and her brother Jordan Steiker (Courting Death: The Supreme Court and Capital Punishment, 2016), Roger Hood and Carolyn Hoyle (cited above), and Franklin Zimring (The Contradictions of American Capital Punishment, 2003). The consensus across their work:

1. Elite political will drives abolition, usually ahead of public opinion. In almost every abolitionist country, polls at the time of abolition showed majority public support for the death penalty. France abolished in 1981 under Mitterrand and Robert Badinter, despite polls showing roughly 60% of the French public in favor of retention. The UK abolished in 1965 with similar polling against. Germany abolished in its 1949 Basic Law with weak public backing. The pattern: elected officials, jurists, and constitutional drafters decided the issue. The public grumbled. Then, within a decade or two, public opinion flipped.

2. Public opinion follows, it doesn't lead. This is the counterintuitive finding. Once a country abolishes, public support for restoration almost always declines over time. The law becomes the baseline, the country's identity reshapes around the new norm, and capital punishment starts to feel like something other countries do. Germany and France are now strongly abolitionist by public opinion. South Africa's polling is more mixed but restoration has never gained political traction.

3. Crime rates don't drive abolition. Several abolitionist countries abolished during periods of rising violent crime. The empirical case on deterrence — does the death penalty reduce murder rates? — is contested but the strongest studies (Donohue and Wolfers, 2005, in the Stanford Law Review, among others) find no reliable deterrent effect beyond that of life imprisonment. But even where abolitionist arguments used this evidence, the decision was rarely driven by it. It was ethical, not empirical.

4. International regimes matter hugely. The European Union's accession requirement (you cannot join with the death penalty on your books) drove abolition in Eastern Europe through the 1990s and 2000s. The Council of Europe's Protocol 6 and Protocol 13 to the ECHR locked it in. The African Charter on Human and Peoples' Rights doesn't require abolition but strongly pushes toward restriction, and regional advocacy groups have moved the continent. The ICCPR's Second Optional Protocol, though non-universal, has been ratified by over 90 states.

This is why abolition feels like a "moral universal" — because it has been built through international instruments that explicitly frame capital punishment as incompatible with universal human dignity. The legal architecture and the ethical argument reinforce each other.

China: Executes more people annually than the rest of the world combined, according to Amnesty International's estimates. Exact figures are classified as state secrets. Estimates range from 1,000 to several thousand per year, though the trend has been downward over the past decade as the Supreme People's Court reclaimed review authority in 2007 and expanded categories of non-violent crimes have been removed from capital punishment (most recently in 2011 and 2015 amendments). Still: the world's largest executioner.

Iran: The second-largest, in absolute numbers and by far the highest per capita among states that publicly report. Executes for drug offenses, "moral" crimes, political dissent, and homosexuality. Public executions still occur.

Saudi Arabia: Beheadings, often publicly, for offenses including drug trafficking, sorcery, and apostasy. Executions have increased in recent years.

United States: An outlier among Western democracies. Executes in the tens annually, concentrated in a small number of states (Texas, Oklahoma, Florida, Alabama, Missouri) — while 23 states have abolished capital punishment and several others have moratoria. The American retention is driven by specific features of American political and legal culture: federalism (each state decides), elected judges and prosecutors (political incentives for "tough on crime" postures), the legacy of lynching and racial terror (which capital punishment continues in legal form, as the disproportionate rates against Black defendants document), and the Supreme Court's role in constitutionalizing the practice through Gregg v. Georgia (1976).

Others in the retentionist camp: Japan (secret executions, usually by hanging, with prisoners notified only minutes before), Singapore (high per-capita rate for drug offenses), Belarus (the only European country that still executes, and a notable Council of Europe holdout), North Korea, Egypt, Bangladesh, Vietnam, and a scattered handful of others.

For the subscriber who wants to understand the scaffolding:

- Universal Declaration of Human Rights (1948): Article 3 ("right to life") is the foundational text. It doesn't explicitly prohibit capital punishment, but subsequent interpretation has increasingly read it as in tension with it.

- International Covenant on Civil and Political Rights (1966): Article 6 permits the death penalty only for the "most serious crimes" and imposes procedural limits. The Second Optional Protocol (1989) requires abolition outright. 91 state parties as of the most recent count.

- European Convention on Human Rights, Protocols 6 (1983) and 13 (2002): Protocol 6 abolishes in peacetime; Protocol 13 extends the ban to all circumstances. Ratified by all Council of Europe states except Russia (expelled 2022) and Belarus (never a member in good standing).

- American Convention on Human Rights, Protocol to Abolish the Death Penalty (1990): Regional instrument for the Americas.

- Arab Charter on Human Rights (2004): Permits the death penalty for the most serious crimes, pursuant to law. No region-wide abolitionist instrument exists in the Arab world.

- African Charter on Human and Peoples' Rights: Permits the death penalty but the African Commission has issued several resolutions urging moratoria.

- UN General Assembly Moratorium Resolutions (biennial since 2007): Non-binding but politically significant. Each cycle has seen increasing support.

Three things.

First: it should recalibrate your sense of moral progress. The arc is real. Capital punishment was universal within living memory. It is not universal now. Humans decided, in scores of countries, that we would stop.

Second: it should complicate your sense of how moral change happens. It's not mostly bottom-up. It's not mostly driven by evidence. It's driven by elite decisions — often contested, often unpopular at the moment of enactment — that shift the baseline and teach the public what a decent society now considers acceptable. The implications for every other moral question you care about are substantial.

Third: it should show you that the holdouts are holdouts. The United States, when it executes a man in Texas, is doing something that most of the world has decided is wrong. Not something merely controversial. Something that the global moral baseline has moved past. That framing matters.

Write down, from memory: 1. Whether your country has the death penalty in law. 2. Whether it has carried out an execution in the past decade. 3. Which international treaty on capital punishment your country has ratified (or failed to ratify). 4. What your own position is, and whether you've ever articulated it.

Most people cannot answer (1) through (3) without looking. That's the educational failure. The moral conversation can't happen if the facts aren't in the air.

- Roger Hood and Carolyn Hoyle, The Death Penalty: A Worldwide Perspective, 5th ed. (Oxford, 2015) — the authoritative empirical text. - Carol S. Steiker and Jordan M. Steiker, Courting Death: The Supreme Court and Capital Punishment (Belknap, 2016) — for the US case. - Franklin Zimring, The Contradictions of American Capital Punishment (Oxford, 2003) — why the US is an outlier. - Amnesty International, Death Sentences and Executions (annual report) — the running empirical count. - S v Makwanyane (Constitutional Court of South Africa, 1995) — read the judgment itself. It is one of the great legal documents of the 20th century and will take you an afternoon.

The abolition of the death penalty is a civilizational "yes" in the making. Not complete. Not inevitable. But visible, directional, and documented.

If you accept the premise of this book — that if every person said yes, we could end hunger and achieve peace — then the death penalty abolition movement is a rehearsal. It shows the yes is possible. It shows how long it takes. It shows who is pulling the lever and who is following.

Your move is to notice you're inside this story. Then to act accordingly.