Democracy for Realists, Part 18 of 19
Economic retrospection mattered, but not in the blanket way that many have supposed. It changed the votes of only a fraction of the electorate, and its impact was sharply channeled by identity groups.
Achen & Bartels devote much of their book to destroying the idea of “retrospective voting” – that is, voters don’t track the actions of politicians to reward or punish them later on. But Achen & Bartels acknowledge that voters do react to good and bad circumstances. The voters simply fail to think about what policies might make a situation better or worse. And however the voters feel about events, their feelings will be shaped by their previous understanding of their own identity.
From the book:
Democracy for Realists, 2016
Why Elections Do Not Produce Responsive Government
By Christopher H. Achen & Larry M. Bartels
Page 314-315
The power of this viewpoint is appallingly illustrated by the most spectacular failure of 20th-century democratic politics. Hitler’s rise to power via the ballot box. The observer in the grip of the folk theory asks, how could Germans have favored those policies? Indeed, anti-Semitism was common in Germany, as it was in most other European countries at the time. But in the view of most scholars, German voters first decided whether they liked Hitler as a potential leader, and then if they did, they adopted his policy views, including his views about “the Jewish question.” They would have agreed with whatever he proposed. “[People] were drawn to anti-Semitism because they were drawn to Nazism, not the other way round” (Allen 1984, 77; for similar judgments, see Hamilton 1982, chap. 13; Fritzsche 1998, passim; Kershaw 2002, chap. 6).
Of course, Hitler’s anti-Semitism was undoubtedly crucial to a minority of voters, but its appeal was limited. As detailed studies of Nazi election propaganda show, “Antisemitism had not been a top priority issue for the Nazis in the last elections of the Weimar Republic” (Gellately 2001, 24). The policy preferences of Hitler’s voters are the wrong place to look for the mainsprings of his rise, just as the platforms of candidates and the ideologies of voters are the wrong place to look for the mainsprings of more typical democratic elections.
Many observers have also invoked retrospective voting to account for Hitler’s electoral success, suggesting that Germans voted for the Nazis because times were hard and because the mainstream political parties had been discredited by prolonged economic distress. In consequence, the Nazi vote, especially in the early elections, had a “catchall” quality, as protest voters in all regions and social classes were swept into Hitler’s coalition (Falter 1990). After Hitler attained power, unemployment declined and Hitler’s popularity rose. Speaking as many Germans did after the war, one woman said of Hitler’s rise to the chancellorship in the midst of protests over unemployment and food shortages, “And naturally all that changed with one blow. My husband immediately got a job, too, and everyone got a job and bread and therefore people were naturally pro-Hitler” (Owings 1993, 177). It is not easy to imagine a clearer example of blind retrospection, nor a more damaging blow to retrospection as a defense of democracy.
Retrospection in hard times affects nearly everyone, and thus it tends to produce a “uniform swing,” a similar shift of all electoral districts against the incumbents and toward the out parties (Butler and Stokes 1974, 140-151). Retrospection in this simple sense, however, does not account well for Hitler’s vote. Many voters in areas dominated by the Socialist and Communist parties resisted him, and Catholics, especially in rural areas, were nearly immune to his appeal (see, for example, Hamilton 1982, chap. 13). Those two identity groups each had very nearly the same percentage of the vote in the 1932 Reichstag elections as they had had in 1924 – a notable anomaly from the standpoint of conventional retrospective explanations. What had collapsed were the old Protestant nationalist parties, the third identity grouping defining German electoral politics since the Wilhelmine period (Lepsius 1966; Burnham 1972; Rohe 1990a; 1990b). The Nazis largely replaced those parties. In the last more or less honest elections of the Weimar era, in 1932, the parties of the right got almost exactly the same total proportion of the vote that they had received in 1924. The difference was that the Nazis were now by far the largest among them, averaging just over a third of the total vote.
Thus, Hitler’s electoral success is a story of identity groups, just as elections virtually always are, but identity groups of a certain kind – not, for example, social classes. Hitler spoke the language of German nationalism in a period of post-war humiliation. Protestant nationalists recognized him as one of their own, though initially not a very respectable exemplar. However, as the horrors of the Depression deepened and made them desperate, they abandoned other parties of the same kind for the sort of strong leadership that Hitler promised. Other parts of the electorate were almost unmoved; Hitler was clearly and decisively anti-socialist and non-Catholic, and thus not one of theirs. Hence economic retrospection mattered, but not in the blanket way that many have supposed. It changed the votes of only a fraction of the electorate, and its impact was sharply channeled by Weimar’s political identity groups.