Massacre memories: German car sales and the EZ Crisis in Greece

Vasiliki Fouka, Hans-Joachim Voth, 23 October 2013



When Angela Merkel visited Athens earlier this year, local protesters upset about austerity measures and alleged German dominance depicted her on placards wearing a Nazi uniform. Cypriots demonstrating against the terms of their country's bailout showed the stars of the EU surrounding a swastika. As the European debt crisis spreads, protests against austerity have become more frequent and often violent (Lynn, 2010 ). In southern Europe the use of Nazi-era symbols has become an integral part of protests against German-led policies in Europe, pointing out what demonstrators see as clear parallels between the present and the past. In times of crisis, EU policy is often seen through the lens of history – in this case, Germany’s history of military aggression and war crimes.

In our working paper, we examine empirically if a confluence between past and present conflict can affect consumer behaviour. We analyse the sales of German cars in Greece after the outbreak of the debt crisis. In times of political conflict, the market share of German brands declines. Particularly in areas where the German armed forces committed massacres during World War II, German car sales fall significantly more than in the rest of Greece.

Conflict after 2009

Forced into a sequence of bailouts as a result of its debt crisis, Greece had to accept harsh expenditure cuts in exchange for funding from the IMF and the EU. Germany took the lead in insisting on austerity.

German newspapers and politicians expressed their contempt over budget fraud and the alleged moral shortcomings of the Greek people. One German newspaper suggested that Greece sell its islands; another showed Aphrodite with a raised middle finger on the cover. In February 2012, the Greek president complained about his country being insulted by the German finance minister.

To measure political conflict, we use a count of newspaper stories on the topic, from the Greek newspaper Kathimerini. We focus on the periods when coverage of public spats ‘spiked.’ Figure 1 shows the evolution of our indicator. There are three periods when conflict was particularly intense:

  • September 2011, when the German finance minister said it was up to Greece to decide if it wanted to stay in the euro.
  • January 2012, when a German plan to take budgetary control away from Greece failed.
  • May 2012, when Greek national elections were held, with one major party campaigning on an anti-bailout platform.

A measure of Google search terms yields a similar picture.

Figure 1. Two measures of German-Greek political conflict

German war crimes in Greece

Fifty civilians for every German soldier – one hundred for every officer. These were the quotas of reprisal shootings to be carried out after partisan attacks, as set down by the German Armed Forces Supreme Command (Mazower 1995). The rules applied to all areas of German occupation, but they were not implemented with any degree of consistency. Instead, depending on the country and the commander in question, reprisal measures varied from modest punishments to – more often – massive attacks on the civilian population unfortunate enough to live in the vicinity of where an attack had taken place. In most cases, they were unconnected with the partisans. German army doctrine did not seek to punish the guilty, but to deter locals from co-operating with irregulars.

In Russia, Yugoslavia, and Greece, German reprisal measures were frequent and harsh. From Crete – where the Wehrmacht committed the first massacres of local villagers who had resisted their advance – reprisals spread to mainland Greece. In one famous incident, the village of Kommeno near the Gulf of Arta came under attack by a German regular army unit in the fall of 1943. Partisans had been observed receiving food in the village; the German reprisals were designed to discourage such support. In total, 317 Greeks died, including 172 women and 13 infants.

As partisan attacks intensified towards the end of the war, villages all over Greece were burnt to the ground. The Greek interior ministry compiled a list of "martyred towns" after the end of the war. To be included, a town or village had to either be entirely destroyed as a result of arson or shelling, have lost 10% of its population in mass executions, or fulfil a combination of these criteria. Seventy-two towns from this list suffered reprisals by the German army, most of them witnessing mass executions. As Figure 2 shows, 'martyred towns' can be found all over Greece.

Figure 2. Martyred towns of Greece

Economic impact

As bad news about the state of the Greek economy accumulated after 2008, car sales slumped. Did the outbreak of the crisis – and public German-Greek conflict in particular – lead to declines in market share for German producers? Since German car producers are heavily focused on the top end of the market, we focus on the ‘VW-category’ of producers of small- and medium-sized cars (such as VW, Toyota, Fiat, etc.).

The overall trend heavily favoured German producers over the period in Europe as a whole – especially because a resurgent VW was taking market share from Toyota. On average, in a month without conflict, German car producers in Greece gained 3.1% market share every 12 months. In months of conflict, this transforms to losses of 0.1%, a marked change.

By looking at German car sales at the prefecture level, we find that there is a pronounced contrast between two types of areas. Those affected by German massacres during World War II saw rapid declines in months of conflict in the German cars' market share – an average of 3.4 percentage points. Areas that didn’t suffer reprisals hardly experienced a change in the German market share compared to ‘normal’ months.

One natural question to ask is how long these effects lasted. We examine the cumulative difference in German cars' market share in massacred and non-massacred prefectures over time, and plot this indicator side-by-side with the share of news articles covering German-Greek conflict. The differential grows with every eruption of political conflict, but there is no quick reversal. By the end of our sample period, the accumulated average difference is almost two standard deviations (equivalent to two percentage points).

Figure 3. Relative decline of German market share (reprisal vs non-reprisal areas) and share of newspaper articles about German-Greek conflict

Did consumers actually switch to other brands, or did they simply postpone purchases? We find that prefectures that saw German massacres in 1941-44 did not only buy fewer German cars – they also bought more cars from other manufacturers. This shows that losses in terms of sales for German car producers during conflict months were not temporary.


Boycotts are amongst the most common forms of political action. On average, almost half of Fortune 50 firms are subject to some kind of boycott at any one point in time. Remarkably, the evidence that boycotts reduce sales is limited. While many US consumers were outraged by France's refusal to support the invasion of Iraq, for example, it is unclear if they actually reduced their purchases of French products (Ashenfelter et al. 2007; Michaels and Zhi 2010; Pandya and Venkatesan 2013). Nor should we be surprised that calls for boycotts typically go unheeded. Consumers "like what they like"; deviating from their ideal consumption basket is costly. While they want others to punish transgressions, they would ideally like to free-ride on the boycott of others (Sen et al. 2001; John and Klein 2003).

In Marcel Proust's A la recherche du temps perdu, the narrator's childhood memories are brought back by the taste of a madeleine. It appears that the bitter taste of the past can have similar effects – revived memories affect the purchases of major consumer goods such as German cars. These findings are related to recent theoretical work that highlights selective, context-dependent memory as a source of behavioural biases (Gennaioli and Shleifer 2010). Our study suggests that where local memory interacts with current conflict, the collective action problem of boycotts can be overcome – memories of past misdeeds can spur consumers to incur the cost that would otherwise stop them from participating in a boycott.


Ashenfelter, Orley, Ciccarella, Stephen and Shatz, Howard J (2007), “French Wine and the US Boycott of 2003: Does Politics Really Affect Commerce?”, NBER Working Paper 13258.

Fouka, Vasiliki and Voth, Hans-Joachim (2013), “Reprisals Remembered: German-Greek Conflict and Car Sales during the euro Crisis”, CEPR Discussion Paper, forthcoming.

Gennaioli, Nicola and Shleifer, Andrei (2010), “What Comes to Mind”, The Quarterly Journal of Economics, 125 (4), 1399–1433.

John, Andrew and Klein, Jill (2003), “The Boycott Puzzle: Consumer Motivations for Purchase Sacrifice”, Management Science, 49 (9), 1196–1209.
Lynn, Matthew (2010). Bust: Greece, the euro and the sovereign debt crisis. London: Wiley.

Mazower, M (1995), Inside Hitler's Greece: The Experience of Occupation, 1941-1944. New Haven and London: Yale University Press
Michaels, Guy and Zhi, Xiaojia (2010), “Freedom Fries”, American Economic Journal: Applied Economics, 2 (3), 256–81.

Pandya, Sonal and Venkatesan, Rajkumar (2013), “French Roast: International Conflict and Consumer Boycotts – Evidence from Supermarket Scanner Data”, mimeo.

Sen, Sankar, Gurhan-Canli, Zeynep and Morwitz, Vicki (2001), “Withholding Consumption: A Social Dilemma Perspective on Consumer Boycotts”, Journal of Consumer Research, 28 (3), 399–417.

Young, Brigitte and Semmler, Willi, “The European Sovereign Debt Crisis: Is Germany to Blame?”, German Politics & Society, 29(1), 2011.

Topics: Europe's nations and regions, Global crisis
Tags: eurozone, Greek debt

PhD candidate in Economics, Universitat Pompeu Fabra

Chair of Development and Emerging Markets at the Economics Department, Zurich University and CEPR Research Fellow