Shanker Satyanath Nico Voigtländer HansJoachim Voth NYU UCLA and NBER UZH and CEPR 26 May 2015 Background Social capital ID: 830204
Download The PPT/PDF document "BOWLING FOR FASCISM: SOCIAL CAPITAL AND..." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
Slide1
BOWLING FOR FASCISM: SOCIAL CAPITAL AND THE RISE OF THE NAZI PARTY IN WEIMAR GERMANY, 1919-33
Shanker Satyanath Nico Voigtländer Hans-Joachim Voth (NYU) (UCLA and NBER) (UZH and CEPR)
26
May 2015
Slide2Background: Social capital
Social capital = a dense network of civic associations “[t]he study of social capital is that of network-based processes that generate beneficial outcomes through norms and trust” (Durlauf and Fafchamps, 2004) Social capital typically associated with benign outcomes: Vigorous democracy (Tocqueville, 1835 and Putnam, 1994,1995)Economic development (Knack and Keefer, 1997) Pro-social behavior (Guiso, Sapienza and Zingales, 2008, 2010)Also acknowledged that social capital can have negative effects, such as social polarization (Durlauf and
Fafchamps
, 2004); crime (Field, 2003)
Autocratic rulers can use social capital as a means of control
entrench elites persistence of “bad” institutions
(
Acemoglu
, Reed, and Robinson, 2013)
This paper: Can social capital also contribute to the fall of existing democracies and the
emergence
of autocracies?
Slide33
Slide4Origins of Totalitarianism:
Mass-society, decline of traditional ties susceptibility for totalitarian doctrineLarge-scale, impersonal, social institutions: Individuals cut off from traditional and communal ties (Ortega y Gasset)Social isolation is key: Appeal to marginal loners on the fringes of society (Arendt, 1958). Stern
(1972): Germany lacked “the kind of voluntary, civic activity that attracted their English and American counterparts.”
Mass
movements – NS, Communists – offer an alternative
identity
weakness of German civic society facilitated rise of the NazisContrast
: Berman
(1997): Weimar Germany
had comparatively high social capital, successfully exploited by NSDAP
4
Slide5This paperConstruct novel dataset for association density in more than 225 towns and cities in the 1920s
Show that there is a robust, strong relationship between density of associations and Nazi Party entry Results equally strong for bowling, singing, and animal breeding clubs etc. Address concerns over endogeneityShow that “dark side” of social capital was particularly strong where regional institutions were weak
5
Slide6Background: Rise of the NSDAPFounded 1919
1923 – attempt to overthrow the government (“Beerhall Putsch”)Party banned in 1924Returns to the polls in 1928, receives 2.6% of the voteIn the late 20s, party organization growing rapidly100,000 members1,400 local chaptersBy 1933, 850,00 members; “Storm Troopers” (SA) strong enough to challenge the government militarily First major electoral success in 1930 (after the party had already built substantial membership base)
Slide7Why party entry mattersMembership expansion via local party cells was a critical pre-condition for subsequent electoral success and the collapse of democracy (Brustein, 1998)
Membership dues key source of financingMembers important for major modes of mobilization (rallies, leafleting, etc.)NS entry rates predict (later) electoral success
Slide8How did the NSDAP exploit local associations? Town chapters (Ortsgruppen
) were responsible for recruiting members, collecting dues, coordinating activities, etc.Koshar (1986) uses the example of Emil Wissner, a salesman in Marburg: Member of a white-collar employee association (from 1921), and active in two gymnastics clubs (from 1904). Joined the party in 1929, and actively used his position to proselytize for the party, and to win new members.
Slide9Social Networks and Party Entry
Fellow association members more trusted than randomly matched acquaintances: makes it easier for Nazi party ‘political entrepreneurs’ to spread the message
One Nazi Party member recounts how he…
“…became acquainted with a colleague of my own age with whom I had frequent conversations. He was a calm, quiet person whom I esteemed very highly. When I found out that he was one of the local leaders of the National Socialist party, my opinion of it as a group of criminals changed completely…”
Slide10NSDAP members were mostly members of ‘ordinary’ associations
From Koshar (1987):
Slide11Data
Collected data on associations from 229 town and city directories, altogether 22,127 associations2.61 associations per 1,000 inhabitants on averageConditions to be in the sample:Contacted all towns larger than 10,000 in 1925 (~550)Surviving directories listing associations in the 1920sFor many towns and cities, this information was lost, destroyed during the war, or it did not exist in the first placeCore measure: “Association density” number of associations per 1,000 inhabitants (Unfortunately, no systematic
data on number of members
)
Matched with sample
Falter-
Brustein
computerized sample
of
38,752 Nazi Party members for 1925-33 (Schneider-Haase, 1991). Identify 9,169 members in the 229 towns and cities
11
Slide1212
12
12
12
Association density
Nazi Party entry
Towns and cities in the sample, by association density and party entry
Note: Full dots = above median; empty dots = below median.
Slide13Slide14Balancedness
Omnibus test:
Test if predicted NSDAP entry (by all controls) is correlated with club density
p-value 0.19 for baseline controls
p-value 0.80 for all controls, conditional on baseline controls
Slide15Baseline results:Nazi Party entry and association density
15
Baseline
Slide16Partial scatter plot, NSDAP entry rate and association density (baseline)
Slide17Baseline results – some robustness
17
Slide18More robustness…
Subsamples by high/low Catholic, blue-collar, city sizeBonding and bridging social capitalMatching estimation (also by location)
18
Slide19NSDAP Party Membership and Success at the Polls
19
Slide20Mediation test: Club density
Party entry Electoral success
Slide21Towards an Interpretation
When do associations matter most?Early vs. late party entry: Stronger results for early entryRole of `Nazi potential’: Associations matter more in towns with lower right-wing votes in early 1920sPanel results: Existing members most ‘useful’ for fostering party growth in locations with high club densityInterpretation:
Local networks particularly useful for new parties that cannot rely on existing members for local recruitment efforts
Further suggestive evidence:
Elections – Counterfactuals
: Weak negative relationship with KPD (communists), none with DNVP (right-wing party) in 1928-33
21
Slide22Understanding the effects of social capital:The role of institutions
A lot of evidence – social capital positive for democracyIn Weimar – contributes to decline and fall of democracyHow do we make sense of the difference?One answer – institutionsStrong, stable institutions align expectationsProvide a positive prior for democracyCan crack down on anti-democratic movements while they’re still small
Slide23Example: Prussia 1919-32 = “The Weimar that works”
Stable Social Democrat coalition (narrow majority)Vigorous defenders of democracy (against right and left)Police stops paramilitary groupsBans political marches (e.g., police shoots on Communist demonstrators on May 1, 1929)Increasing pressure from central government starting 1930. Prussian government deposed in a putsch in July 1932
Slide24Proxy for government stability (state-level):
Three indicators over the period 1918-July 32 (ending with the Prussian coup d’etat)the percentage of time that the longest-serving government was in officethe percentage of time that the longest-serving party was in office (possibly in different coalitions)the percentage of time that a state was governed by at least one party from the “Weimar coalition” (SPD, DDP,
Zentrum
)
Compute first
principal
component
Prussia in the upper third, but some states equally or more stable
Slide25The role of stable government – preview
Low stability, outside Prussia
(N=58)
Prussia
(N=119)
High stability, outside Prussia (N=48)
Slide26Slide27The net effect of club density on NSDAP entry
Slide28Towards causalityIV
Additional checks“Plausibly exogenous”Selection on unobservables Altonji et al.
28
Slide29IV strategy
1860s membership in gymnast associations1861 participation in the Nuremberg Singing Festival283 singing associations6,000-20,000 participants[Friedrich Nietzsche among them]Exclusion restriction19C nationalism vs. national socialism (“spirit of 1848”)Turn towards the apolitical after 1850Evidence from Democratic congress in Berlin, 1848
Slide30IV results
Slide31ConclusionsSocial capital associated with many beneficial outcomes
We showed: dense networks of clubs associated with rise of the Nazi party, and thus contributing to the collapse of Weimar Germany’s democracyDoes not depend on the type of associationImportant interaction with institutions: Social capital can be exploited to undermine democracy when institutions are weak
Slide32Tocqueville knew it all along… “The most natural privilege of man… is that of combining his exertions with those of his fellow creatures and of acting in common with them. The right of association therefore appears to me almost as inalienable in its nature as the right of personal liberty. … Nevertheless, if
the liberty of association is only a source of advantage and prosperity to some nations, it may be perverted or carried to excess by others, and from an element of life may be changed into a cause of destruction.” [italics added]
Slide33BACKUP
33
Slide34The basic result
Cumulative NSDAP membership, by association densityNote: Each data point reflects the cumulative NSDAP entry rate (per 1,000 inhabitants), starting in 1925 and averaged across the cities with above- and below- median association density. The data are described in Section 3. NSDAP entries are from the Berlin-Minneapolis sample (Schneider-Haase 1991); starting in 1930, we correct aggregate entry rates for a change in sampling methodology, as described in Appendix C.
Total entries per 1,000
Below-median club
denisty
: 27.7
Above-median club density: 35.2
Slide35Partial scatter – exclude 90pctile of club density
35
Slide36Historians’ judgment
“…the movement may have ‘fed’ on preexisting membership structures and clusters which the single members helped channel into the reestablished Nazi party. Ties with other ‘bourgeois’ associations of Germany’s civil society may well have enhanced this process by extending the reach of single members for recruitment purposes.” (Anheier 2003) “Weimar's rich associational life provided a critical training ground for eventual Nazi cadres and a base from which the NSDAP could launch its Machtergreifung (seizure of power).” (Berman 2007)
Slide37Example: Coburg
Slide38Veterans of
5
th
infantry
Regiment
Bamberg
Soccer
And athletics
Slide39Bad Langensalza[member count not listed]
Slide40Associations in the sample
40
Slide41Members per capita vs. Clubs p.c.
41
Membership data for sports and gymnastics clubs members available for Prussian towns in 1927
Slide42Elections – Counterfactuals
42
Slide43Early and late Nazi Party entries, by locality
Note: The x-axis plots average rates of Nazi Party entry (per 1,000 inhabitants) in each city over the period 1925-28, and the y-axis over the period 1929-33. Data are described in Section 3.
Slide44Slide45Slide46Slide47Location-specific ‘sociability’? Worker associations as counterfactual
Evidence for “general sociability”, but not for an effect on Nazi party entryWorking class associations did not act as gateways to the Nazi movement – "infection" apparently required a minimum degree of ideological compatibility
Slide48Slide49Slide50Civic vs. Military clubs
Slide51“Bridging” vs. “Bonding” social capital
Slide52Bonding and bridging social capital
Slide53Matching estimation and geographic location
Slide54Quantile regressions
Slide55Quantile regression – graph
Quantile regression graph
Note: The figure shows the effect of a unit increase in association density on Nazi Party entry rates, by quantile of the dependent variable. The shaded area reflects the 95% confidence interval of the quantile regressions. The figure is derived for our main specification, with the three baseline controls: share of Catholics, ln(population), and the of share blue collar workers, all measured in 1925. Absolute coefficient sizes are plotted. For standardized beta coefficients, see Table 13 in the paper.
Slide56Annually standardized entry rates
56
Slide57Plausible exogeneity – Results following Conley, Hansen, and Rossi (2012)
Comparison: Reduced-form coefficient on IV is 0.227.
Slide58Selection on unobservables - Altonji ratios
Slide59Association delegates to Democratic Congress in Berlin, 1848