The 1931 banking crisis and the Nazi vote - the engine's cleanest economic-shock-times-culture result
eventMoney & Finance · Media & Managed Opposition
One bank failed in 1931, the towns that banked with it got poorer than everywhere else, and then they voted Nazi harder than everywhere else. The bank next door failed too and changed nothing.
Who they are
Danatbank, a major German bank led by a prominent Jewish manager and named directly in Nazi propaganda; and Dresdner Bank, which collapsed in the same crisis.
What they do
Economists reconstructed which German towns did business with which bank before the crash. In the towns tied to Danatbank, incomes fell 7.8 percentage points further than the average 14-point drop across German cities — and the Nazi vote rose an extra 2.9 percentage points between September 1930 and July 1932.
How it works
The comparison is what makes it strong. Dresdner Bank's collapse had an effect on Nazi voting that the authors describe as statistically and economically zero. The difference between the two banks was not their size; it was that one of them was already a target of antisemitic campaigning. And the voting effect showed up most in towns where antisemitism was already well established. Two causes working together, not one. The work is published in a major finance journal; the exact percentages quoted here come from the authors' own 2020 manuscript, which was downloaded and read.
Why it matters
This site's Weimar page calls the collapse a textbook case of too many educated people chasing too few good jobs. That may well be part of it. This is a second, measured mechanism operating on the same voters in the same years, and neither one cancels the other out.
The engine's record — word for word
[Report #202] **In plain terms:** when a particular bank failed, the towns that banked with it got poorer than everywhere else and then voted Nazi harder than everywhere else - but only that bank, and only where antisemitism was already established. Two causes, interacting. Neither one alone does it.
THE NUMBERS, verbatim from the authors' 17 November 2020 manuscript, downloaded and read this pass: 'We establish that municipalities more exposed to collapsing Danatbank suffered sharper economic declines. Their incomes during the crisis fell by 7.8 percentage points (p.p.) more than the average 14 p.p. decline across cities.' And: 'in locations exposed to Danatbank there was a clear upward shift in voting for the Nazis that added up to 2.9 p.p. to the party’s gains between September 1930 and July 1932.' The 14 p.p. baseline is carried deliberately: a 7.8-point gap against a 14-point average decline is a different claim from 7.8 points below a national average, and the research pass dropped the baseline.
THE PART THAT MUST TRAVEL WITH THE NUMBERS. The effect is not generic financial distress. Dresdner Bank also collapsed in 1931 and the paper finds its effect on Nazi voting 'statistically and economically zero'. Only Danatbank moved votes - a bank 'led by a prominent Jewish manager and targeted by anti-Semitic Nazi propaganda' - and the effect concentrated 'especially in cities with deep-seated anti-Semitism'. The authors' own framing is 'a synergy between cultural and economic factors'. This is therefore not a pure substrate finding and must not be filed as one.
PUBLICATION STATUS, stated precisely because the two differ. The work is peer-reviewed: Journal of Finance 77(6), December 2022, pp. 3339-3372, DOI 10.1111/jofi.13166, under the title 'Financial Crises and Political Radicalization: How Failing Banks Paved Hitler’s Path to Power'. That published abstract states direction, not magnitude - 'Nazi votes surged in locations more affected by its failure', 'exacerbated in cities with a history of anti-Semitism', and that after the seizure of power 'both pogroms and deportations were more frequent in places affected by the banking crisis'. The 2.9 and 7.8 figures above are from the 2020 manuscript, not from the published abstract; the published PDF could not be reached this pass. An earlier VoxEU column by the same authors states 2.5 p.p. rather than 2.9.
WHAT THIS DOES TO CANON'S WEIMAR RECORD: weimar_elite_overprod opens 'Turchin textbook case.' The edge nsdap -> weimar_elite_overprod already hedges it as 'the credentialed-overproduction reading, held as a partial mechanism per Report #201'. This record is what that hedge was waiting for - a competing, measured, peer-reviewed mechanism operating on the same electorate in the same years. Neither displaces the other and no apex is moved. Held; name no holder.
Follow the trail
Walk this on the live map →