Blame games: Malattribution and the Madrid train bombings

  • Spring ∕  Summer 2022
Blame games: Malattribution and the Madrid train bombings

Nina Miller is a PhD student in security studies and international relations in MIT’s Department of Political Science. Her research interests include military innovation, ambiguity and decision-making, and strategic stability. She is a recipient of a Kenan Sahin Presidential Graduate Fellowship at MIT.

SPRING/SUMMER 22 :: précis Student Feature :: Nina Miller
Nina Miller
June 29, 2022

On the morning of March 11, 2004, ten bombs detonated on four commuter trains in Madrid, killing 192 and injuring 1,400.[1] The next day, two million people gathered in Madrid’s Cibeles plaza in a silent protest. When the president and royal family arrived, the silence was finally broken by a whistle and shouts of “Who was it? Who was it?”[2] Two days later, on the morning of the national election, the Spanish people learned that despite official claims that the Basque terror group ETA was responsible, the attack had actually been the work of Al-Qaeda affiliates.[3]

Why do leaders sometimes blame the wrong group for attacking their state? After an act of political violence, the first question most people ask is who is responsible. Blame can be a powerful political force that spurs anger and mobilizes people. When leaders shape the public narrative about blame, it can allow them to deflect criticism and pursue policy goals. What happened in Madrid is an example of what I call malattribution, or deliberate misattribution. Malattribution is a puzzling and potentially risky choice for leaders to make, because blaming the wrong group can make it more difficult to punish the true perpetrators. Leaders also risk getting caught in their deception, thereby decreasing their international credibility and domestic trust and support.

What is attribution?

Leaders often have to decide how to publicly respond to high-profile acts of violence, such as bombings or assassinations. The public is generally intolerant of uncertainty and seeks cognitive closure about what happened, who did it, and why.[4] Given this demand for an explanation, leaders need to decide what secret information they share, when to update the public on progress in the investigation, whether or not the leader personally weighs in on the situation, and the degree of confidence with which they state their conclusions. Elite cues can shape how the public thinks and feels about a complicated and ambiguous situation.[5] By framing the public narrative about the event, how a leader responds can affect support for the leader and their policies and perceptions about the leader’s legitimacy and credibility.

Attribution is a key aspect of the investigation and public narrative. If attribution were a purely technical process, a state’s primary focus would be to establish reasonable confidence in the perpetrator and communicate this to their citizens. However, attribution is also a political choice about how to deliberately and publicly communicate secret information about who is responsible for an attack. Even if two states experienced the same attack, possessed the same evidence, and used the same investigative apparatus, we would expect to get different attribution outcomes. Leaders affect attribution outcomes by resourcing and shaping the investigation, helping to draw conclusions, and then communicating those conclusions to third parties and the public.[6]

Strategies of malattribution

There is a greater variety of deliberate attribution choices than whether or not to blame a bad actor. Leaders have private information about who is likely or unlikely to be responsible for an attack. However, they may selectively share or distort this private information in order to overstate the likelihood a group is—or is not—involved. In other words, leaders may find it politically advantageous to deviate from the body of evidence available to them in order to publicly claim that an innocent group is involved or that a guilty group is not. I use the term malattribution to refer to this deliberate public distortion of private information to overstate or understate confidence that a group is responsible for a negative event.

I identify four different strategies for leaders to shape the public narrative in the aftermath of political violence: true attribution, linking, exoneration, and substitution. Leaders can publicly blame the actor who is most likely responsible for the attack—true attribution. They can also claim that an innocent actor aided the true culprit, thereby falsely linking them to the attack. Alternatively, leaders can deflect blame from the true perpetrators in an attempt to exonerate them. The Madrid case is an example of substitution, when leaders focus blame on an innocent actor and exonerate a guilty one.

I argue that leaders select a malattribution strategy based on the magnitude of the threat to the regime and policy goals. Public perceptions about what has happened and why can either bolster regimes, providing support to pursue domestic and foreign policy goals, or they can be destabilizing and existentially threatening.[7]

In general, I argue that leaders publicly under-state their true beliefs that a guilty actor is involved when they anticipate that telling the truth will threaten their political survival. Political leaders prioritize threats that could remove them from power, because survival in office is necessary for them to pursue any other policy goals.[8] In democracies, leaders will worry that this opposition will translate into electoral losses. Responsiveness to public opinion is not limited to democracies, however. Autocrats also face threats to their survival and depend on a narrow, loyal portion of the public to stay in power.[9] Regimes face a political threat when the true perpetrator is a foreign group that the leader’s own policies have antagonized or otherwise failed to contain. Additionally, leaders face a threat to their legitimacy when the true perpetrator is a domestic group with which the leader or their base has supported or participated.

I expect leaders to over-state their true beliefs that an innocent actor is involved when they anticipate that the deception will help them achieve pre-existing policy goals. Scapegoating can deflect criticism of the leader by mobilizing supporters against a common outgroup, like a domestic ethnic minority or a foreign adversary.[10] This logic will be especially strong when the scapegoated group is perceived as a competent but cold actor that is both capable and willing to harm the ingroup.[11] Blaming a political opponent can improve one’s own electoral competitiveness by creating perceptions that the opponent was involved or responsible for the attack. Shifting blame can also generate “rally-around-the-flag” effects with increased support for military action or certain legislation.

Malattribution in Madrid

Hours after the attack in Madrid, Minister of the Interior Ángel Acebes blamed Euskadi ta Askatasuna (ETA or “Basque Homeland and Freedom”), militant separatists from the northern territory of Basque Country.[12] When asked if it could possibly be an actor like Al-Qaeda, Acebes reiterated the government’s confidence that “we have no doubt [that it was ETA], nor do I believe that the overwhelming majority of Spaniards do, when we have seen that little over a week ago [ETA] were looking for this exact attack.”[13] Other members of the conservative Partido Popular (PP, or “People’s Party”) repeated the attribution over the next 56 hours, including PP presidential candidate Mariano Rajoy.[14] Hours before polls for the national elections opened on March 14, Acebes finally admitted Al-Qaeda had credibly claimed responsibility and that five individuals affiliated with Al-Qaeda had been arrested in Madrid.[15]

There is only suggestive evidence about when Spanish leadership knew Al-Qaeda was responsible and to what extent the incorrect attribution was deliberate. Hours after the attacks, Prime Minister José María Aznar called then-director of El País Jesús Ceberio to emphasize multiple times that there was “absolute certainty” about ETA’s guilt.[16] Journalists at the state-run news agency EFE later said they had information about Al-Qaeda’s involvement the morning of the attacks, and accused director Miguel Platón of imposing “a regime of manipulation and censorship” to favor the PP ahead of the election.[17] Spanish diplomats pressured the United Nations Security Council to adopt Resolution 1530, which “[c]ondemns in the strongest terms the bomb attacks in Madrid, Spain, perpetrated by the terrorist group ETA.”[18] Spanish diplomats insisted they acted in good faith, but the resolution was adopted quickly and with uncharacteristic certainty about the bombers’ identity. Members of the Security Council later complained they “were utilized for political maneuvering” and felt the council had been “hijacked.”[19]

The conduct of the investigation provides additional evidence that PP elites may have deliberately shifted blame away from Al-Qaeda and towards ETA. Members of the police reported speaking with Minister of the Interior Ángel Acebes and other People’s Party politicians about evidence pointing to Al-Qaeda, yet minutes after these conversations, Acebes held a press conference to declare ETA was responsible.[20] The head of the bomb disposal unit Tedex Juan Jesús Sánchez Manzano insists that he knew with certainty that Al-Qaeda was responsible the afternoon of March 11.[21] Then-director of the National Intelligence Center Jorge Dezcallar has described how the government asked him to publicly declare on March 13 that ETA was still under investigation—despite the arrest of suspects in connection with Al-Qaeda—and felt the PP was trying to use him for partisan advantage.[22]

What explains PP decision-making in the aftermath of the Madrid train bombings? The imminent election increased the salience of public opinion and created an imminent threat for Aznar and other PP leadership. Spain had joined the US-led coalition in the 2003 invasion of Iraq, despite a full 90 percent of Spaniards opposing the war.[23] Given the domestic unpopularity of the war, conservative leadership may have anticipated angry voters would blame them for entering Iraq and failing to prevent the Madrid attack. Subsequent analysis has suggested that the Partido Popular would have won the 2004 elections if not for the attacks.[24] In surveys from after the election, a full thirty percent of Spaniards said the attack affected their vote.[25]

Spanish leadership also had good reasons to prefer blaming ETA. The Partido Popular had a hardline anti-ETA platform,[26] and Aznar himself had survived an ETA assassination attempt in 1995.[27] As journalist Giles Tremlett noted at the time, Aznar and Rajoy would benefit if ETA were responsible, because “both support a ‘no negotiating’ stance on ETA, and the group has seen its operational capacity seriously reduced by police action during Mr Aznar's eight years as prime minister.”[28] PP leadership could also have expected their base of conservative voters to believe their claims about ETA’s guilt and support hardline policies. A February 2004 poll found that 43.2 percent of Spaniards thought ETA’s terrorism was one of Spain’s top three problems.[29]


The weaponization and manipulation of blame is an issue central to the study of politics. Leaders make strategic decisions about attribution in a variety of contexts, including cyber competition, terrorism, financial crises, false flags, and accidents. Studying malattribution offers insight into key junctions in domestic politics and international relations. How and why do leaders use secrecy, collusion, and deception to achieve policy goals? Why are comparatively small acts of violence sometimes the catalyst for long-lasting and costly wars? How can elite cues unify a country after political violence—and when do these events become divisive, polarizing symbols?

Aznar and other PP politicians appear to have gambled that substitution would pay off in an electoral victory. Although any single factor likely would not have motivated malattribution, Spanish leaders faced significant political threats—including the overwhelming domestic opposition to the Iraq war and imminent election—as well as an opportunity to pursue hardline policies against ETA due to widespread uncertainty about what had actually happened.

However, insisting on ETA’s guilt provoked a public backlash that may have cost the Partido Popular an election. The Madrid attacks were a turning point in Spanish politics and foreign policy. Immediately after taking office, the new socialist administration withdrew from Iraq. US officials raised concerns that the withdrawal might motivate future attacks if Al-Qaeda drew “the wrong lesson” about the success of terror in Madrid.[30] Soon after the election, members of the People’s Party began claiming that ETA and Al-Qaeda had acted in collaboration. This apparent strategy of linking ETA and Al-Qaeda has cast a long, conspiratorial shadow on Spanish politics and sowed mistrust in official explanations.[31] Nearly two decades after the Madrid attacks, the attribution question remains contentious.

[1] “Tres días de tragedia en Madrid,” El País, 2014, sec. 11 Marzo - Décimo Aniversario,; José Manuel Romero, “Cuatro atentados simultáneos causan una matanza en trenes de Madrid,” El País, March 12, 2004, sec. España,

[2] Pablo Ordaz, “¿Quién ha sido?,” El País, March 13, 2004, sec. España,

[3] Elsa Granda, “La verdad de Acebes, paso a paso,” El País, July 27, 2004, sec. España,; “Tres días de tragedia en Madrid.”

[4] Christopher M. Federico, Agnieszka Golec, and Jessica L. Dial, “The Relationship Between the Need for Closure and Support for Military Action Against Iraq: Moderating Effects of National Attachment,” Personality and Social Psychology Bulletin 31, no. 5 (May 1, 2005): 621–32,; Marta Marchlewska, Aleksandra Cichocka, and Małgorzata Kossowska, “Addicted to Answers: Need for Cognitive Closure and the Endorsement of Conspiracy Beliefs,” European Journal of Social Psychology 48, no. 2 (2018): 109–17,

[5] Alexandra Guisinger and Elizabeth N. Saunders, “Mapping the Boundaries of Elite Cues: How Elites Shape Mass Opinion across International Issues,” International Studies Quarterly 61, no. 2 (June 1, 2017): 425–41,; Charlotte Koot et al., “Facilitation of Attitude Formation through Communication: How Perceived Source Expertise Enhances the Ability to Achieve Cognitive Closure about Complex Environmental Topics,” Journal of Applied Social Psychology 46, no. 11 (2016): 627–40,

[6] Thomas Rid and Ben Buchanan, “Attributing Cyber Attacks,” Journal of Strategic Studies 38, no. 1–2 (2015): 4–37.

[7] Arturas Rozenas and Denis Stukal, “How Autocrats Manipulate Economic News: Evidence from Russia’s State-Controlled Television,” The Journal of Politics 81, no. 3 (July 2019): 982–96,

[8] Bruce Bueno de Mesquita et al., The Logic of Political Survival (The MIT Press, 2003),

[9] Bueno de Mesquita et al.

[10] Ala’ Alrababa’h and Lisa Blaydes, “Authoritarian Media and Diversionary Threats: Lessons from 30 Years of Syrian State Discourse,” Political Science Research and Methods 9, no. 4 (October 2021): 693–708,

[11] Susan T. Fiske, “Envy up, Scorn down: How Comparison Divides Us,” American Psychologist 65, no. 8 (2010): 698–706,; Peter Glick, “Choice of Scapegoats,” in On the Nature of Prejudice (John Wiley & Sons, Ltd, 2005), 244–61,; Susan T. Fiske et al., “A Model of (Often Mixed) Stereotype Content: Competence and Warmth Respectively Follow from Perceived Status and Competition,” Journal of Personality and Social Psychology 82, no. 6 (2002): 878–902,

[12] Paddy Woodworth, “Boys Become Giant-Killers: The Making of ETA,” in Dirty War, Clean Hands (Yale University Press, 2001), 33–34,; Paul Preston, The Spanish Holocaust: Inquisition and Extermination in Twentieth-Century Spain, 1st edition (W. W. Norton & Company, 2013), 428–40.

[13] Granda, Elsa. “La verdad de Acebes, paso a paso.” El País. July 27, 2004, sec. España.

[14] Granda; “Tres días de tragedia en Madrid.”

[15] Granda, “La verdad de Acebes, paso a paso”; “Tres días de tragedia en Madrid.”

[16] Jesús Ceberio, “A propósito de mentiras,” El País, March 26, 2004, sec. España,; Miguel Catalán, “Prensa, verdad y terrorismo: la lección política del 14-M,” El Argonauta español. Revue bilingue, franco-espagnole, d’histoire moderne et contemporaine consacrée à l’étude de la presse espagnole de ses origines à nos jours (XVIIe-XXIe siècles), no. 2 (January 15, 2005),

[17] Inter Press Service, “Silencing the Truth About the Attacks in Spain,” Antiwar.Com (blog), March 19, 2004,

[18] “Resolution 1530 (2004)” (United Nations Security Council, March 11, 2004),

[19] Elaine Sciolino, “Bombings in Madrid: Intelligence; Many in Europe Suspect Spain Misled Them About Attackers,” The New York Times, March 16, 2004, sec. World, One ambassador’s spokesman admitted that despite their frustration, “we cannot prove that at the time Spain didn't trust its information.” See also Therese O’Donnell, “Naming and Shaming: The Sorry Tale of Security Council Resolution 1530 (2004),” European Journal of International Law 17, no. 5 (November 1, 2006): 945–68,

[20] Granda, “La verdad de Acebes, paso a paso.”

[21] The explosive material was white—rather than the red Titadyn that ETA used—and far more potent. Sánchez also says that Aznar’s government asked him to lie to the congressional investigation into 11-M, saying that Tedex had identified the explosives as Titadyn. Patricia Ortega Dolz, “‘El Gobierno de Aznar me pidió que asumiera su mentira sobre el 11-M,’” El País, March 11, 2019, sec. Política,

[22] José Gómez, 11M: Terror in Madrid (Netflix, 2022). Dezcallar says the Secretary of State of Communication called him six times on March 13 insisting “on behalf of the president” that he publicly blame ETA. See “Jorge Dezcallar: ‘Lo garantizo. Yo no sabía que iban a atentar en Madrid el 11-M,’” El Mundo, October 4, 2015, sec. Cronica,

[23] 900 Spanish soldiers moved from Kuwait into southern Iraq soon after the U.S. invasion to participate in humanitarian assistance and subsequent combat roles. See Stephen A. Carney, “Allied Participation in Operation Iraqi Freedom” (Washington, D.C.: Center of Military History: United States Army, 2011), 6; 17; 110,; J. L. Lorente, “Nueve de cada diez españoles se oponen a la guerra, pero la mayoría cree que Irak es un peligro mundial,” ABC, March 27, 2003, sec. Internacional,

[24] Looking at voters who cast their vote before the bombing, José Montalvo finds that the bombings did have an important effect on vote choice: José G. Montalvo, “Voting After the Bombings: A Natural Experiment on the Effect of Terrorist Attacks on Democratic Elections,” The Review of Economics and Statistics 93, no. 4 (2011): 1146–54. See also Javier Noya, “Del 11-M al 14-M: estrategia yihadista, elecciones generales y opinión pública” (Real Instituto Elcano de Estudios Internacionales y Estratégicos, 2004),

[25] Noya, “Del 11-M al 14-M: estrategia yihadista, elecciones generales y opinión pública,” 2.

[26] The PP 2004 platform boasted PP success in jailing ETA members at home, as well as international cooperation to crack down on ETA including the United States designating the group as terrorists and French cooperation to jail or extradite ETA members. “Avanzamos Juntos: Elecciones Generales 2004 Programa de Gobierno,” 305; 409, accessed March 16, 2022,

[27] Jesús Duva, “ETA intenta asesinar a Aznar con un coche bomba,” El País, April 19, 1995, sec. España,

[28] Giles Tremlett, “Eta Still Main Suspect despite Denials,” The Guardian, March 13, 2004, sec. World news,

[29] Islamist terror “did not appear in the results.” From Nuria Quintana Paz, “Televisión Pública y 11-M: La Información Sobre Terrorismo En Campaña Electoral,” Comunicação & Cultura, no. 4 (2007): 99.

[30] Marlise Simons, “Spanish Leader Pulling Troops From Iraq,” The New York Times, April 18, 2004, sec. World,

[31] Jorge A. Rodríguez, “La teoría conspirativa del 11-M, capítulo a capítulo,” El País, September 29, 2006, sec. España,; Irene Castro, “El PP alimentó la teoría de la conspiración del 11M con cientos de preguntas parlamentarias,”, March 9, 2014, sec. Política,