Peter Selb and Simon Munzert electoral turnout in the 1928-1930 election pair,indi- (or larger decreases)in NSDAP vote shares were to be cating that Hitler's early appearances had,on average expected.One interpretation of this result would be a demobilizing effect in exposure areas.A look at the that,for infrastructural reasons,Hitler's appearances other outcomes suggests that this demobilizing effect were scheduled in more populous urban areas where tended to harm the KPD but left the Nazi (population) the NSDAP tended to fare worse electorally for social vote share unaffected.This result supports Childers structural and programmatic reasons (Heberle 1978; and Weiss's (1990)observation that a concomitant Thurner,Klima,and Kuchenhoff 2015) feature of Nazi campaign events was concerted vio lence to intimidate political opponents and their sup- porters.This specific empirical pattern disappears in ROBUSTNESS later elections,perhaps indicating that Goebbels'at- tempt to align the paramilitary branch of the NSDAP Absence of evidence is no evidence of absence.In (Sturmabteilung,SA)to the quasilegal course adopted this section,we will play devil's advocate and ask if earlier was,at least in part,successful.Although es- there are reasons to suspect that actual campaign ef- timated exposure effects on NSDAP vote shares are fects went undetected with our data and research de- mostly zero,a negative effect occurs with the 1930-32 sign.The discussion is organized around three potential election pair-just when the Nazis took their greatest problems:SUTVA violations due to spatial spillovers leap forward electorally! and multiple versions of exposure,violation of the The 1932 presidential runoff election deserves spe- causal transience assumption with respect to previous cial mention.Incumbent president Hindenburg had Hitler appearances,and the limitations of our data and emerged as the clear front-runner from the first elec- inferential approach to identify the effects of dynamic tion round on March 13,1932,garnering 49.5%of the campaign strategies vote.By comparison,Hitler won 30.1%of the vote.and communist leader Ernst Thalmann secured 13.5%.All other candidates combined received less than 7%17 and Spatial Spillovers and Multiple Versions of did not stand again in the decisive second round.Hin- Exposure denburg,who did not bother to enter the electoral fray One serious drawback of our study is the nonavail- himself (Pyta 2007,475),issued an emergency decree ability of exposure data.Like other spatial ecological that limited the campaign period preceding the runoff studies,we have to make do with spatial proximity to on April 10 to just the six days prior to the election- events as a surrogate for exposure.Thus far,the spec- a decision that prompted Hitler to utilize the airplane ification of the exposure variable has arbitrarily fixed for the first time as a means of transportation for cam- the geographic reach of Hitler appearances to a radius paigning.The novel mode of transportation afforded of ten kilometers around venues.If the actual reach him 21 campaign appearances spread over the whole of events was wider than that,campaign effects might country during this short period,with a maximum of have spilled over to neighboring areas that served as five appearances per day.Even though Hitler even- potential comparison units in estimating(in this case. tually lost the election to Hindenburg,he still gained underestimating)effects(see Rubin 1980).In the previ- more than two million additional votes from the first ous analysis,we drew no-matching zones around expo- 5795.801g to the second round.Our analysis suggests that some of sure areas to avoid this problem.As an alternative,we the vote gains actually came from his campaign appear- now vary the specifications of exposure areas from five ances.The finding that an unusually short and intense to 50 kilometers around event locations for the 1928- campaign had significant effects is in line with results 1930 municipal-level data,which provide the highest from previous research,which indicates that campaign spatial resolution.We likewise vary the range of the no- effects run out quickly over time(e.g.,Watt,Mazza,and matching zones from five to 50 kilometers.The results Snyder 1993)-for example.due to the interference of plotted in Figure 5 indicate that initially positive ef- offsetting campaign messages.To more systematically fects of visits on NSDAP vote shares as well as initially account for potential effect decay,we replicated our negative effects on communist vote shares and turnout analysis,but this time only considering campaign ap- tend to fade out with increasing spatial distance.While pearances that took place within two,four,eight,and these findings do not immediately suggest an optimal twelve weeks before an election.The results are,how- cutoff value for specifying exposure areas,our hunch ever,inconclusive.The details are reported in Tables I7 is that a radius of 10 km plus an additional 10 km to 121 in the Appendix. no-matching zone is a safe choice.The potential for A final methodological remark,the matching ap- even broader diffusion effects (e.g.,through national proach obviously makes a difference.Difference-in- media)seems limited given the Nazis'limited access to differences estimates of exposure effects on Nazi vote public media;its mostly negative stance toward Hitler shares are consistently more negative when applied before 1933 (Adena et al.2015):the low circulation to the unmatched samples.This indicates that Hitler's figures of the Nazis'own national paper,Volkischer campaign trail usually stopped where smaller increases Beobachter(Layton 1970);and the geographically frag mented press landscape at that time(Fuhrer 2008). 17 The Social Democrats and the catholic Center Party did not nom Apart from non-interference between units,SUTVA inate separate candidates.Rather,they supported Hindenburgr to also requires that there be only a single version of treat- prevent the election of Adolf Hitler. ment(Rubin 1980).This assumption is unlikely to hold 1060Peter Selb and Simon Munzert electoral turnout in the 1928–1930 election pair, indi￾cating that Hitler’s early appearances had, on average, a demobilizing effect in exposure areas. A look at the other outcomes suggests that this demobilizing effect tended to harm the KPD but left the Nazi (population) vote share unaffected. This result supports Childers and Weiss’s (1990) observation that a concomitant feature of Nazi campaign events was concerted vio￾lence to intimidate political opponents and their sup￾porters. This specific empirical pattern disappears in later elections, perhaps indicating that Goebbels’ at￾tempt to align the paramilitary branch of the NSDAP (Sturmabteilung, SA) to the quasilegal course adopted earlier was, at least in part, successful. Although es￾timated exposure effects on NSDAP vote shares are mostly zero, a negative effect occurs with the 1930–32 election pair—just when the Nazis took their greatest leap forward electorally! The 1932 presidential runoff election deserves spe￾cial mention. Incumbent president Hindenburg had emerged as the clear front-runner from the first elec￾tion round on March 13, 1932, garnering 49.5% of the vote. By comparison, Hitler won 30.1% of the vote, and communist leader Ernst Thälmann secured 13.5%. All other candidates combined received less than 7%17 and did not stand again in the decisive second round. Hin￾denburg, who did not bother to enter the electoral fray himself (Pyta 2007, 475), issued an emergency decree that limited the campaign period preceding the runoff on April 10 to just the six days prior to the election— a decision that prompted Hitler to utilize the airplane for the first time as a means of transportation for cam￾paigning. The novel mode of transportation afforded him 21 campaign appearances spread over the whole country during this short period, with a maximum of five appearances per day. Even though Hitler even￾tually lost the election to Hindenburg, he still gained more than two million additional votes from the first to the second round.Our analysis suggests that some of the vote gains actually came from his campaign appear￾ances. The finding that an unusually short and intense campaign had significant effects is in line with results from previous research, which indicates that campaign effects run out quickly over time (e.g.,Watt,Mazza, and Snyder 1993)—for example, due to the interference of offsetting campaign messages. To more systematically account for potential effect decay, we replicated our analysis, but this time only considering campaign ap￾pearances that took place within two, four, eight, and twelve weeks before an election. The results are, how￾ever, inconclusive. The details are reported in Tables I7 to I21 in the Appendix. A final methodological remark, the matching ap￾proach obviously makes a difference. Difference-in￾differences estimates of exposure effects on Nazi vote shares are consistently more negative when applied to the unmatched samples. This indicates that Hitler’s campaign trail usually stopped where smaller increases 17 The Social Democrats and the catholic Center Party did not nom￾inate separate candidates. Rather, they supported Hindenburgr to prevent the election of Adolf Hitler. (or larger decreases) in NSDAP vote shares were to be expected. One interpretation of this result would be that, for infrastructural reasons, Hitler’s appearances were scheduled in more populous urban areas where the NSDAP tended to fare worse electorally for social structural and programmatic reasons (Heberle 1978; Thurner, Klima, and Küchenhoff 2015). ROBUSTNESS Absence of evidence is no evidence of absence. In this section, we will play devil’s advocate and ask if there are reasons to suspect that actual campaign ef￾fects went undetected with our data and research de￾sign.The discussion is organized around three potential problems: SUTVA violations due to spatial spillovers and multiple versions of exposure, violation of the causal transience assumption with respect to previous Hitler appearances, and the limitations of our data and inferential approach to identify the effects of dynamic campaign strategies. Spatial Spillovers and Multiple Versions of Exposure One serious drawback of our study is the nonavail￾ability of exposure data. Like other spatial ecological studies, we have to make do with spatial proximity to events as a surrogate for exposure. Thus far, the spec￾ification of the exposure variable has arbitrarily fixed the geographic reach of Hitler appearances to a radius of ten kilometers around venues. If the actual reach of events was wider than that, campaign effects might have spilled over to neighboring areas that served as potential comparison units in estimating (in this case, underestimating) effects (see Rubin 1980). In the previ￾ous analysis, we drew no-matching zones around expo￾sure areas to avoid this problem. As an alternative, we now vary the specifications of exposure areas from five to 50 kilometers around event locations for the 1928– 1930 municipal-level data, which provide the highest spatial resolution.We likewise vary the range of the no￾matching zones from five to 50 kilometers. The results plotted in Figure 5 indicate that initially positive ef￾fects of visits on NSDAP vote shares as well as initially negative effects on communist vote shares and turnout tend to fade out with increasing spatial distance. While these findings do not immediately suggest an optimal cutoff value for specifying exposure areas, our hunch is that a radius of 10 km plus an additional 10 km no-matching zone is a safe choice. The potential for even broader diffusion effects (e.g., through national media) seems limited given the Nazis’ limited access to public media; its mostly negative stance toward Hitler before 1933 (Adena et al. 2015); the low circulation figures of the Nazis’ own national paper, Völkischer Beobachter(Layton 1970); and the geographically frag￾mented press landscape at that time (Führer 2008). Apart from non-interference between units, SUTVA also requires that there be only a single version of treat￾ment (Rubin 1980). This assumption is unlikely to hold 1060 Downloaded from https://www.cambridge.org/core. Shanghai JiaoTong University, on 26 Oct 2018 at 03:56:49, subject to the Cambridge Core terms of use, available at https://www.cambridge.org/core/terms. https://doi.org/10.1017/S0003055418000424