International Influences and Democratic Regression in Interwar Europe: Disentangling the Impact of Power Politics and Demonstration Effects

Abstract

Scholars have convincingly argued that theoretical frameworks that combine international influences and domestic factors are needed to understand political regime developments. We argue that exogenous shifts in the balance of power between great powers (‘power politics’) spark demonstration effects. These, in turn, are filtered into the domestic political system of smaller states via changes in political polarization – but with the effects being conditional on the domestic vulnerability of democracy. To assess this framework we turn to interwar Europe, where the international order changed from undergirding democracy to facilitating autocratic rule. An analysis of three countries (Poland, Austria and Denmark), backed by a more general glance at the remaining interwar cases, shows that the interaction between demonstration effects, pressure from great powers and the domestic resilience of democracy offers substantial leverage in accounting for patterns of democratic regression.

Recent studies have demonstrated the importance of integrating international and domestic factors to make sense of political regime developments (Boix 2011; Levitsky and Way 2010; Weyland 2009, 2010). However, research on the link between international influences and domestic regime change suffers from two shortcomings. First, many studies are preoccupied with the post-Cold War period, which is characterized by a general pro-democratic pressure. To fully appreciate the effects of the interplay between external and internal conditions, it is necessary to analyse periods where the international environment takes on different values (see Boix 2011; Gunitsky 2014; Owen 2010; Weyland 2009, 2010, 2012). Second, there is disagreement about how different kinds of international influences interact and which types of influence matter most for democratic development and decay. The general dividing line runs between scholars who emphasize the importance of demonstration effects (e.g. Brinks and Coppedge 2006; Huntington 1991; Weyland 2009, 2010) and scholars who stress the importance of direct influences from great powers based on the balance of power between democratic and autocratic countries (e.g. Boix 2011; Carr 1946; Levitsky and Way 2010; Owen 2010).

In this article, we relate the two kinds of international influences to each other and turn to interwar Europe, a context in which democracy was first backed by a liberal hegemony in international affairs but was subsequently put under stress by one facilitating autocracy. The main aim is to deepen our understanding of how changes in the international environment produce pro-democratic and pro-authoritarian empirical manifestations across different domestic contexts. The secondary aim of the article is to shed new light on democratic developments in interwar Europe where a rapid initial expansion in the number of democracies was soon replaced by a negative trend of reversals until the end of the Second World War (Berg-Schlosser 2015: 338). A voluminous scholarship deals with this topic, but the interactions between international and domestic factors have not received much attention (see, e.g. Berg-Schlosser and Mitchell 2002; Luebbert 1991; Mann 2004; Rueschemeyer et al. 1992).

The Argument

To understand how external influences filter down to affect domestic political developments, we need to integrate the two external dimensions of demonstration effects and direct external pressure from great powers (what we term ‘power politics’). We define demonstration effects as innovations adopted or events happening in one site that inspire and embolden actors in other societies to press for or shore up against similar developments; power politics we understand as direct foreign pressure wielded by especially great powers on target states.

We combine these two dimensions in a punctuated framework where exogenous shifts in the balance of power among great powers spark demonstration effects. In brief, when democratic powers predominate, pro-democratic demonstration effects proliferate and democratization flourishes; when autocratic powers preponderate, anti-democratic demonstration effects abound and democratic regressions dominate.

Our argument resembles Gunitsky’s (2014) notion that abrupt changes in the international system (‘hegemonic shocks’) tend to produce waves of political regime change. But Gunitsky solely focuses on direct interventions and more indirect influences from great powers; he does not take an explicit stand on the relative importance of power politics and demonstration effects or the exact way they interrelate, and he does not address how different domestic contexts in target countries may condition the impact of the external stimuli. We go further in an attempt to understand how different kinds of international influences interact and which kinds of external influences are more effective.

We stress that the normal mode of external stimuli is one of demonstration effects. But we add that the magnitude and the plenitude of these effects are largely determined by developments in the balance of power in the international system. For democracies and autocracies alike, anti-democratic demonstration effects serve to increase polarization between existing groups in the political system, and to stimulate, and radicalize, the formation of new political movements (Weyland 2012), thereby influencing political regime developments. The more frequent and the more powerful the demonstration effects of the international environment are, the more likely it is that opposition groups and rulers respond to cues from abroad and decide to act to either challenge or protect the status quo. For instance, anti-democratic demonstration effects inspire, embolden and further radicalize anti-systemic political movements and tendencies, and they alarm established pro-democratic forces in democracies and ruling groups in less democratic states, pushing them to stand firmer and increasingly resort to non-democratic means to counter real or perceived challenges, even if only for shorter periods. Thus, as democratic great powers are challenged internationally and become less inclined to protect and promote democracy abroad, anti-democratic demonstration effects are likely to become stronger and to spur democratic regression by increasing political polarization.

Next, we also go beyond Gunitsky (2014) – and most previous research on external influences – by arguing that political regime developments are affected in different ways across different domestic contexts (see also Boix 2011); that is, the impact of anti-democratic demonstration effects on democratic regression is conditioned by the domestic vulnerability of democracy. As noted, anti-democratic external stimuli mostly influence regime developments by triggering latent political conflicts or reinforcing manifest ones. In countries where democracy is backed by favourable background conditions, such as a long democratic legacy and socioeconomic development, anti-democratic demonstration effects have little bite as moderate political parties are better at and more committed to shoring themselves up against enemies of democracy within and beyond their own ranks. Yet in vulnerable democracies, anti-democratic demonstration effects increase political extremism, which can either make governments more prone to resort to repression to counter challenges or lead to outright political regime change. Indeed, even where democracy has already broken down, pro-authoritarian demonstration effects might further political extremism and thereby be conducive to increased state repression. To sum up, shifts in the balance of power that strengthen autocracies at the expense of democracies trigger anti-democratic demonstration effects in a general sense, but actual emulation is likely to be much more pervasive in countries with unfavourable domestic conditions for democracy because extremists will have an easier time mobilizing against the already weak democratic centre once anti-democratic demonstration effects fan the flames of political polarization.

We examine these theoretical claims by confronting them with the empirical variation in interwar Europe. We do so in an analysis of three countries (Poland, Austria and Denmark), backed by more general glances at the remaining interwar cases.

An Integrated Theoretical Framework

The starting point of our theoretical perspective is Boix’s (2011) notion that the international system works as a scope condition for how domestic political struggles play out, and therefore also for when democratic regime change is likely to take place. Sometimes the international order is conducive to democracy (democratic order), sometimes it facilitates and undergirds autocracy (anti-democratic order), and sometimes it does not matter much (neutral order). Boix (2011) claims that the shape of the international system depends on two factors: the profile of the dominant great powers (democratic or non-democratic) and the way international alliances are structured (along or across political ideologies). If democratic great powers are dominant on the global scene and if they primarily form alliances with other democracies (as in the post-Cold War order), they are expected to protect both strong and weak democracies and put pressure on autocracies to democratize. If, on the other hand, democratic great powers are constrained by non-democratic great powers and if alliances follow ideological lines – as during the Cold War – we should expect less favourable conditions for democracy as the democratic great powers, fearing to lose allies to the enemy, will be less inclined to uphold the relatively underdeveloped and therefore more fragile democracies and will not try to push autocrats out of power (see also Westad 2005).

In contrast, Weyland (2009: 408–16; 2010: 1151–4; 2012: 29) argues that targeted actions of great powers are in most instances not very important for explaining waves of regime challenges and changes and instead emphasizes the power of demonstration effects and the bounded rationality of domestic actors. Major political developments often leap from one country to another as domestic actors take cues from events happening outside their countries. When popular uprisings or regime changes unfold in focal cases (most often great powers or cases emblematic of certain regime types), opposition elites, incumbent elites, and the population at large tend to overestimate the likelihood that the event can be successfully replicated in their own setting. As a consequence, both efforts to emulate and to shore up against emulation are likely to be pursued not only in cases where successful replication is a real possibility but also in unpropitious cases where failure is the most likely outcome.

These two perspectives thus offer starkly competing expectations with regard to which of the two dimensions of the international order are likely to have the greatest impact and for how long they matter. While one argues that external effects will be more or less constant when the international order takes a particular value, the other holds that the most important international influences emanate from intensive shocks rather than constant pressure.

The premise of our theoretical framework is that both perspectives have explanatory purchase and that they need not be incompatible. As Gunitsky’s (2014) theory of hegemonic shocks demonstrates, they can be combined into a punctuated account where shocks unleashed by changes in the balance of power between great powers provide the inflection points between different periods. In contrast to Gunitsky, we only expect direct pressure from great powers on target states to have a marginal direct effect on domestic developments. Moreover, pace Boix, democratic great powers are in any case unlikely to systematically intervene to defend other democracies as great powers generally navigate in terms of the national interest (see Owen 2010). Thus, the balance of power between democracies and autocracies is expected to mostly have an indirect effect on domestic political developments via the demonstration effects they trigger – that is, the impact of power politics can be said to follow inadvertently through its inspiration of domestic groups.

Table 1 shows how these arguments can be combined to construct a fine-grained typology of the external dimension that incorporates both the active and the passive elements. Six types of international orders can accordingly be distinguished: a pure and a challenged form of democratic hegemony, contested orders with either pro-democratic or pro-authoritarian shocks, and a pure and a challenged variant of authoritarian hegemony. We have shaded the three types that are relevant for capturing developments in interwar Europe.

Table 1 Typology of International Orders

The outlined approach has at least two advantages. Firstly, it offers a less rigid understanding of international influences than seeing them as either pro-authoritarian or pro-democratic. Secondly, this multifaceted view of international influences explicitly incorporates the dimension of power politics in an attempt to more clearly specify when demonstration effects are apt to have more of an impact.

In a nutshell, pro-democratic demonstration effects are more likely to be pervasive and more effective in periods characterized by democratic hegemony and vice versa with anti-democratic demonstration effects. Furthermore, sudden changes in great power configurations can themselves be viewed as powerful demonstration effects. These shifts in the balance of power – triggered either by the ascendancy of a new great power or by regime change in established great powers – are exogenous to our theory about external influences.

How, more specifically, are external influences imported into the domestic scene and how do they affect regime change? External impulses, be it in the form of demonstration effects or outright interference of other states, trigger and reinforce political conflicts by strengthening, inspiring, emboldening and alarming established and emerging domestic actors (see Owen 2010: 39–40). The consequent increase in polarization and conflict can either produce outright regime change in the form of democratic breakdown or, more frequently, gradual democratic regressions, ¹ which are the product of the government resorting to increased repression to counter (sometimes even pre-empt) challenges from extremists. Finally, even after democratic breakdown, increased polarization can trigger hikes in state repression, including further curtailment of civil liberties, disbandment of parliaments and undermining of the rule of law.

However, we do not expect international influences to affect all countries alike. It is important to understand that some political systems will be relatively immune to demonstration effects whereas others will be much more receptive. ² Arguably, the way in which each of the six international orders produces anti-democratic domestic manifestations in existing democracies is conditional on key determinants of domestic vulnerability. In countries with propitious conditions for democracy, pro-democratic political actors are able to maintain democracy even in the face of intense anti-democratic impulses. Countries with unpropitious conditions are much more prone to be affected by anti-democratic demonstration effects.

Contextualizing the Framework

Interwar Europe displays interesting variation not only on patterns of democratic regression, but also on the other variables of interest, domestic democratic fragility (across space) and the intensity and character of the international influences (across time). The universe of European democracies in the interwar period can be subdivided into three clusters, demarcated by different levels of domestic vulnerability of democracy, and the international environment in this period changed significantly, from a pro-democratic liberal hegemony to a challenged liberal hegemony and then to a contested order characterized by a proliferation of pro-authoritarian demonstration effects.

Regarding the domestic strength of democracy, we have clustered countries on the basis of the four domestic, structural factors included in Berg-Schlosser’s (2002: 282) ‘Analytic Map of Europe’: social heterogeneity, feudalism, modernization and pre-war democratic legacy (see also Berg-Schlosser 2015: 344–5). ³ These conditions also figure prominently in other previous explanations of interwar regime change (Luebbert 1991; Mann 2004; Rueschemeyer et al. 1992). We use dichotomous variables for strong landlords and ethnic heterogeneity from Møller et al. (2015) to operationalize the first two conditions. Next, we have constructed a dichotomous variable capturing whether a country had, at least, a minimalist democratic regime before the end of the First World War (Skaaning et al. 2015). To capture modernization in a broad sense, we use the factor scores from an unrotated factor analysis, including data on GDP/cap (Bolt and van Zanden 2014), literacy (Vanhanen 2003), life expectancy, and urbanization (CLIO-Infra 2015). ⁴ The cluster analysis (Ward linkage) produces the following three cluster solution:

1. 1 Albania, Bulgaria, Greece, Portugal, Romania, Yugoslavia.

2. 2 Austria, Czechoslovakia, Estonia, Germany, Hungary, Italy, Latvia, Lithuania, Poland, Spain.

3. 3 Belgium, Denmark, Finland, France, Ireland, the Netherlands, Norway, Sweden, Switzerland, the UK.

The third cluster corresponds to the ‘North-Western’ category identified by Mann (2004), comprising the Low Countries, the British Isles, Scandinavia, France and Switzerland. As Mann (2004: ch. 2) points out, these countries had the potential to resist anti-democratic international influences, whether via demonstration effects or power politics. Against this, Mann places an encompassing category of inherently fragile democracies (see also Luebbert 1991; Payne 1996: 129–31). The cluster analysis divides this second category of fragile democracies into a group with very low domestic democratic resilience (the first cluster) and one with medium democratic resilience (the second cluster).

The way international influences produced domestic political manifestation – and hence affected democracy – in each of these three groups in the interwar years can be understood with reference to the three different international orders shaded in Table 1: A period of democratic hegemony (1919–21), a period where democratic hegemony was increasingly challenged (1922–32) and a pro-autocratic period (1933–8).

During the period of democratic hegemony, the democratic great powers – the US, Britain and France – were so powerful that democratic demonstration effects proliferated. It was not that the democratic powers actively attempted to spread democracy in this period via outside pressure or interference. ⁵ Nonetheless, domestic elites tended to anticipate that the victorious democratic great powers ‘would provide material benefits to those who jumped on the democratic bandwagon’ (Peceny 2010: 3). Furthermore, due to the domestic fragility of many of the new democracies, this period was one of almost perpetual political instability in large swathes of Europe. This instability was further fuelled by the Russian Revolution in 1917, a pivotal event that in 1919 inspired the creation of the non-democratic Hungarian Soviet Republic, the Bavarian Soviet Republic and several uprisings elsewhere. However, in 1919, as world revolution failed, these anti-democratic demonstration effects ebbed and the model of the day was a democratic regime form, drawing inspiration from the powerful victors of the First World War, which collectively made up the by far most powerful pole in the international order.

The more lasting consequence of the Russian Revolution was to alert and unite rightist forces elsewhere in Europe, inspiring them to take power to rule out the communist scenario, whether this scenario was realistic or not (D’Agostino 2012: 165). However, according to our theoretical framework, such a mobilization of the undemocratic right only became important after shifts in the balance of power had provided actual non-leftist authoritarian alternatives to democracy in great powers. During the period of democratic hegemony (1919–22), we should thus be able to identify some mobilization of the right as a response to the perceived threat of a communist takeover. But, in general, these should be cancelled out by the demonstration effects – conducive to political moderation and to strengthening the democratic centre of the political spectrum – unleashed by the pro-democratic international order.

This brings us to the second period – where democratic hegemony became challenged – inaugurated by the US withdrawal back into isolationism, the British preoccupation with its overseas colonies and dominions, the revival of classic interstate relations following the Treaty of Rapallo in 1922 (which meant that the undemocratic USSR was no longer a pariah) and Mussolini’s takeover of power in Italy in 1922 (D’Agostino 2012). Even though Mussolini’s Italy might not pose much of a geopolitical challenge to the democratic great powers, his ascendancy further polarized politics in other European democracies, as a powerful example of a successful and progressive takeover from the right was now added to the already present scenario of a communist revolution. Due to this partial shift in the world order, we expect pro-democratic demonstration effects to be on the wane and anti-democratic equivalents on the rise, in particular among the more fragile cases in Eastern, Central and Southern Europe.

However, the most important inflection point of interwar Europe can be dated to Hitler’s ascent to power in 1933 and the subsequent rapprochement between fascist Italy and Nazi Germany, culminating in the ‘Rome–Berlin Axis’ of November 1936. It is rather obvious that the onset of the Great Depression in 1929 paved the way for this shift as it did much to discredit the Western democracies, which had until then favoured free-market economies that now ‘lost much of their appeal’ (Davies 2006: 140) and as it generally worked as a trigger of democratic breakdown in the interwar years (Møller et al. 2015). But, as already mentioned, shifts in the balance of power between democracies and autocracies are exogenous to our theoretical model. What matters is the dating of the definitive shift in the balance of power. Here, we note that it was only after 1933 that the upper hand in European international affairs definitively oscillated from the democratic to the autocratic camp (D’Agostino 2012; Gunitsky 2014). The Hitlerite takeover and its successful project of resurrecting Germany, reinforced by Mussolini’s increasingly repressive and assertive stance, emboldened rightist forces to take and consolidate power with anti-democratic means. Hence, during the period characterized by a contested order (1933–8), we expect that rightist mobilization increased even further due to a cascade of pro-authoritarian demonstration effects. This facilitated an increasing number of democratic breakdowns in fragile European countries and it was also conducive to the intensification of repression in already authoritarian regimes.

Empirical Analysis

Since we expect the international effects to be conditional on domestic factors, we need variation along the two dimensions described above: the international order (as represented by the three periods) and the level of domestic vulnerability. We refrain from including a case from the most fragile group (Cluster 1) where democracy was basically doomed from the beginning. Instead, we select Austria and Poland from the medium fragility category (Cluster 2) and Denmark from among the resilient ‘North-Western’ countries (Cluster 3).

Due to its almost impeccable scores on all the domestic structural conditions identified above, we expect the Danish case to be immune to anti-democratic demonstration effects throughout the interwar period, even after the inflection point of 1933 when such demonstration effects cascaded. From among the more fragile cases in Cluster 2, the selection of Poland and Austria serves to enlist relevant variation in two respects: they differ with respect to the timing of democratic breakdown (1920s versus 1930s), and they differ on several background conditions even though they are placed in the same cluster. Most obviously, Austria was more modernized and more ethno-linguistically homogeneous than Poland. We expect both cases to show evidence of non-democratic demonstration effects in the 1920s, with a significant increase after 1933. However, we expect democracy to be more resilient in Austria than in Poland, meaning that the non-democratic demonstration effects of the 1920s have less bite in the former than the latter case.

Each case is analysed throughout the period 1919–38, thereby securing the needed temporal variation. In addition, the case studies are supplemented with a wider comparative glance at the remaining European countries.

A major challenge exists in spelling out the mechanisms of demonstration effects and power politics, respectively, in a way that can be investigated empirically at case level over time and that enables us to assess if neither, one or both sets of mechanisms identified in the theoretical framework operate. Recall that we conceive of demonstration effects as innovations adopted or events happening in one site that inspire and embolden actors in other societies to press for or shore themselves up against similar developments and power politics as direct external pressure from great powers on target states. Below, we formulate two sets of observable implications with respect to both dimensions. This procedure makes it possible to scrutinize whether our integrated framework offers more leverage than each of its constituent parts.

The implications of the manifestation of demonstration effects are that domestic anti-democratic forces tried to copy the example of fascist Italy and/or Nazi Germany regarding organization, ideology and means to take and consolidate government power; that leaders of domestic anti-democratic groups (either publicly or in private correspondence) invoked the example of fascist Italy and/or Nazi Germany when deciding about and carrying out democratic regressions; and that domestic leaders stepped up repression against extremist forces with reference to the danger of the fascist or communist takeovers that had occurred elsewhere. The implications of manifestations of influence due to power politics are that fascist Italy and/or Nazi Germany directly supported anti-democratic forces through political, economic or military means; that fascist Italy and/or Nazi Germany directly influenced autocratic takeovers through political, economic or military means; and that France and/or Britain bolstered democratic governments vis-à-vis extremist forces or prevented autocratic takeovers through political, economic or military means.

In each of the three cases, we describe patterns of democratic regression (be it in the form of decreasing electoral integrity or increasing power concentration or state repression) and assess the extent to which these regressions were associated with either type of external stimulus. In principle, the detection of demonstration effects and power politics is decoupled from the dependent variable, democratic regression. That is, both external dimensions may be important factors for understanding different actors’ intentions and capabilities without necessarily leading to observable changes in regime developments.

Poland

Reflecting the dominant political model prevailing in the aftermath of the First World War, Poland established a democratic regime in 1919. However, the Polish state that was reborn following the First World War provided an inauspicious domestic setting for democracy, with low levels of development, ethnic divisions, a strong class of landlords and no tradition of parliamentary rule (Holzer 2003: 339; Rothschild 1974: 52). It is therefore unsurprising that the elections of 1919 inaugurated a period characterized by a teeming of different parties and frequent changes of coalitions and cabinets (Rothschild 1974: 46). This parliamentary instability was finally brought to an end with Marshal Josef Pilsudski’s coup d’état in mid-May 1926.

To what extent were this coup or the preceding developments inspired by external examples or furthered by external pressure? The Polish right was fascinated by Mussolini and eulogized his coup, whereas the left had initially condemned the 1922 March to Rome and its consequences. Nonetheless, even Pilsudski’s supporters, who were mostly placed on the non-communist left, soon softened their attitude to the example set by Mussolini and saw the potential for Pilsudski to play an analogous role in Poland. Pilsudski himself (and his closest associates) did not buy into the fascist ideology (Borejsza 1980: 355–7). However, Pilsudski was clearly inspired by Mussolini’s tactics and thought he could copy his actions and gain power without bloodshed, which was important to him. As he initiated his March to Warsaw, ‘[h]e knew that Mussolini had been able to complete his march on Rome despite a glaring imbalance in military power, and like Il Duce he expected to overwhelm government forces with moral authority, not bullets’ (Hetherington 2012: 560).

When perusing what has been written about Polish democracy and the 1926 coup d’état, it is clear that there was no direct outside meddling in political affairs, including attempts by Britain and France to defend democracy (Bideleux and Jeffries 1998: 464). In fact, the British were quick to support Pilsudski as they hoped he would be able to create political stability, strengthen the military and come to terms with Germany so that Poland could defend itself against the Soviet Union. The French were more worried because they feared losing an important ally, not because Polish democracy had broken down (Rothschild 1966: 297–9).

Despite the absence of external criticism and counter-measures, the coup was not initially followed by harsh repression. There existed ‘a general national craving to demonstrate that the reborn Poland was, despite the coup, sufficiently mature to emulate successfully the Western model of constitutional parliamentary government’ (Rothschild 1974: 61). In other words, democracy had not been totally discredited. Pilsudski created an instance of what we would today term ‘electoral authoritarianism’, headed by a series of technocratic governments, designed to place national interests over partisan ones. Unlike the fascist regimes, Pilsudski imposed political passivity rather than mobilization upon the population, and he expressly avoided making analogies to fascism (Holzer 2003: 350; Rothschild 1974: 59-61).

However, as fascism and dictatorial rule became fashionable in other European countries in the 1930s, repression was gradually stepped up in Poland. Even before the 1926 coup, the communists faced some restrictions to their freedom of the press but the grip was further tightened in 1926, 1930, 1935 and 1938 (Bates 1998). A new constitution implemented in 1935 increased the powers of the executive and turned parliament into a consultative body with ‘people’s’ representatives rather than party representatives. Meanwhile, the freedom allowed to agrarian and socialist groups was reduced to a minimum (Kochanowski 2001). Moreover, openly fascist trends and a general radicalization of all major political camps came to characterize Polish political life in the second half of the 1930s (see Kochanowski 2001: 84–5; Rothschild 1974: 71; Seton-Watson 1945: 165–8). At one end of the spectrum, the rightist opposition to Pilsudski organized around the National Conservative ‘Endeks’ developed in an anti-Semitic and fascist direction. Already in 1934, a splinter party, the so-called National Radical Party, presented a Nazi-inspired programme but the Endeks themselves were not far behind.

One of the reasons for Pilsudski’s more repressive stance in the early 1930s was to counter these challenges to his right. But a more repressive system was also supported by a radicalization within the ruling elite, which culminated in 1937 as pro-Pilsudski groups organized the National Unity (OZN) movement, deliberately modelling it on totalitarian parties of fascist countries. These developments were not based on demands from neighbouring great powers – that is, Hitlerite Germany to the west and Stalinist Soviet Union to the east. Also, the democratic great powers continued to support the Polish government for the sake of power balancing and cared little about political repression in Poland.

Austria

Following the breakdown of the Habsburg Empire, a large majority of Austrians wished to unite with neighbouring Germany (Gerlich and Campbell 2000: 40), but France and Britain vetoed this out of fear that it would strengthen post-war Germany unduly. During much of the 1920s, Austria was thus placed under a form of tutelage of the Western powers. Already in November 1918, when Austria was on the verge of a severe food crisis in the aftermath of the First World War, Otto Bauer, the social democratic acting foreign minister, explained to the Staatsrat that help could only come from the West, meaning that Austrian foreign policy had to have a Western orientation. Bauer further explained that the democratic great powers were equally against monarchical and Bolshevist dictatorships – and that to make life easier for itself, Austria had to become a democracy (Hanisch 2011: 156). Nonetheless, the impetus for democracy primarily came from within. Both among elites and in the general population there was strong consensus to follow the political model of the day (Brauneder 2000: 89–93).

When Austria entered the interwar period, its level of socioeconomic development was high, its population was ethnically homogeneous; but it did not have a strong democratic legacy to draw on as government had never been responsible to the parliament and the country contained a class of strong and politically influential landlords. After an initial coalition between the dominant parties of the right (the Christian Social Party) and the left (the Social Democratic Party), a bipolar or even bisected system developed after the elections in October 1920 as the conservative forces continuously barred the Social Democrats from national government. At the same time, the municipal government of Vienna – tellingly titled ‘Red Vienna’ – was a Social Democratic stronghold.

Mussolini’s takeover in 1922 immediately served to radicalize the domestic political scene in neighbouring Austria (Gerlich and Campbell 2000: 45). In 1923, the Austrian Social Democratic Party created its own paramilitary organization (Republikanischer Schutzbund). As a response, the Heimwehr, a loosely organized movement of former soldiers established to secure public order and protect the Austrian borders after the war, was reorganized as a counterweight by conservative forces. The Heimwehr ‘was anti-parliamentary and fascist, taking Mussolini’s fascist Italy as a reference model, but at the same time it was also anti-Nazi’ (Gerlich and Campbell 2000: 45). More generally, the Austrian forces of the right were divided into two camps: the Christian Social Party with close connections to Italian fascism, and the Austrian National Socialists looking to Nazi Germany. Most members of the Heimwehr belonged in the former camp and they were propped up by successive conservative chancellors.

The result was a tense political atmosphere where armed Heimwehr troops often clashed with troops from the Schutzbund. In 1927, things came to a head during the so-called ‘July Revolt’. The left initiated massive protests when right-wing radicals were acquitted of having killed a man and a child. The Social Democrats called a general strike on 15 July, aiming to bring down the conservative government. The Ministry of Justice was torched and police assisted by Heimwehr troops dispersed the demonstrating workers, killing 84 protesters and wounding more than 600 people. The conservative forces justified the actions of the police with reference to the danger of a ‘red plot’ in the form of a Bolshevik revolution. But the explosive situation – an instance of extreme political polarization – did not cause the Social Democrats to call out the Schutzbund in an attempt to seize power or the right wing to use the incident as an excuse for a coup.

Democratic breakdown only came after Hitler’s ascendancy to the post of German chancellor in January 1933. In March, the Austrian chancellor, Englebert Dollfuss, took advantage of the resignation of the president and two vice presidents of the Nationalrat to close parliament and replace democracy with a so-called ‘Austrofascist’ regime. Along the line of Italian fascism, Dollfuss established the Vaterländische Front, and in February 1934 the government set about destroying the Social Democratic Party completely. Fighting between Schutzbund and Heimwehr forces (the latter backed by the police and army) left almost 1,000 dead. The Social Democratic forces were defeated and the repression of civil liberties intensified.

The external influences and reactions varied between the neighbouring great powers. As early as 1928, Mussolini began to supply arms and money to the Heimwehr and pleaded for a clampdown on the Austrian Social Democrats, who still controlled Vienna (Burgwyn 1997: 49). Dollfuss repeatedly turned to Il Duce for guidance, which was readily supplied (Maderthaner and Maier 2004). Dollfuss and Mussolini carried on an intensive correspondence, in which they openly discussed the possibility of a complete takeover and repression of the Social Democrats. However, Germany also interfered in 1933–4 by inciting the Austrian Nazis to begin a terrorist campaign against the Dollfuss dictatorship, culminating in the July 1934 Putsch. Nazi forces occupied the Chancellery and assassinated Dollfuss. Several days of fighting broke out in the border provinces where the Austrian Nazis were strong. Germany actively aided the July Putsch from across the border in Bavaria. However, Mussolini was not interested in having Nazism defeat fascism in Austria. He responded by mobilizing the Italian army across the Brenner and threatening an invasion. Germany backed down, and Dollfuss’ successor, Kurt von Schuschnigg, used the non-Nazi part of the Heimwehr to suppress the Nazis. This state of affairs continued until the German–Italian rapprochement sealed by the Pact of Steel in 1936; after that it was only a matter of time before the Anschluss would become a reality.

Throughout the dramatic events of 1933–4, Britain largely remained aloof. Dollfuss’s coup was not even met with official criticism since interference in Austrian affairs was considered unwise and was believed to play into the hands of more extremist forces. So when the French government, where especially the socialist ministers argued in favour of attempting to reinstall democracy in Austria, expressed disapproval of the authoritarian tendencies, the British tried to make the French tone down their criticism and demands. British support for Dollfuss faded as a consequence of the violent suppression of the Social Democrats in February 1934, but throughout the events of the early 1930s none of the democratic great powers took any concrete actions to protect Austrian democracy (Bertolaso 1995).

Denmark

Denmark had a strong democratic legacy to rely on when it entered the interwar period, in addition to a high level of socioeconomic development, an ethno-linguistically homogeneous population and an absence of a strong class of landlords. Throughout the 1920s and 1930s, Denmark figures as a case of extraordinary domestic vigour with regard to rejecting the authoritarian trends of the day. Though we do witness several examples of existing and newly formed political groups finding inspiration and gathering self-confidence from the anti-democratic demonstration effects that abounded at the time, public support for extremist movements remained marginal and the leaders of the four major parties repeatedly showed a commitment to democracy, moderation and inclusive policies. All of them fought political extremism, whether it appeared in their own ranks or elsewhere (Lauridsen 2003). Thus, Denmark appears as a perfect specimen of the ‘North-Western’ trajectory, where democracy was preserved – and often even strengthened – in the face of the anti-democratic sentiments that flourished elsewhere in Europe. The Danish case lends strong support to one of our main theoretical arguments: that the domestic context is key to understanding where anti-democratic external influences have bite.

The only time up until the German occupation in 1940 when the survival of Danish democracy was genuinely threatened was in the early days of democratic hegemony when pro-authoritarian demonstration effects were absent. During the so-called Easter Crisis (Påskekrisen) in March and April 1920, King Christian X made use of a long-dormant clause in the constitution which allowed him to dismiss the Social-Liberal government. Not only did the king ignore political conventions at the time, he transgressed the principle of parliamentarism by acting against the will of the majority in the Danish parliament. The reason was that he, and the two major parties on the right, the Liberals and the Conservatives, felt that larger parts of formerly Danish territory should be returned by Germany (despite a German majority in these areas). Though these dramatic events unfolded at the peak of democratic hegemony, we have come across no evidence that the democratic great powers tried to prevent or counter this soft coup d’état. On the contrary, French diplomats, with the intent of further weakening Germany, eagerly lobbied the Danish king to take action (Kaarsted 1968).

The Social-Liberal government, which was backed by the Social Democrats, was outraged but nevertheless decided to counter the coup with non-violent demonstrations and soft diplomacy. Characteristic of the Danish case, the crisis was resolved peacefully within a week. New elections were called, and the parties that had supported the king’s actions, the Liberals and the Conservatives, emerged as victors (Kaarsted 1968). Surprisingly, the conflict between the two groups in parliament did not escalate, power changed hands peacefully, the new government showed restraint and willingness to compromise, and democracy was quickly reinstated and even strengthened. From this point on, a modus vivendi was established, according to which all important political players exhibited a commitment to democracy, moderation and inclusive policies.

This acceptance of democracy as the only game in town also explains why the ‘fascist’ demonstration effects in the period of challenged democratic hegemony during the 1920s do not seem to have played much of a role. We see no significant radicalization of any of the major parties during this decade, and though anti-systemic movements like the Society for Recovery (Genrejseforeningen) and the Movement for Autonomy (Selvstyrebevægelsen) ⁶ were clearly inspired by the ideas and the charismatic style of Mussolini, no extremist organizations gained parliamentary representation or even popular backing beyond a few thousand supporters. Furthermore, no party-based paramilitary organizations were created, and political violence was absent (Lauridsen 2003: 69–72; Lidegaard 2003: 259–62). Had political disagreements between the established parties of the Danish parliament been more intense, or had Danish society performed worse in terms of socioeconomic determinants, it is likely that the first wave of pro-authoritarian demonstration effects could have caused more political conflict in parliament and resonated more with Danish society in general, potentially triggering democratic regressions.

More serious challenges emerged in the 1930s. As Hitler’s National Socialists won a solid foothold in the German Reichstag in 1930 and later gained complete power, the external threat to Danish democracy increased. Ever since the decisive defeat in the Second Schleswig War in 1864, Danish political elites had been careful not to provoke or anger Germany, and Hitler’s rise to power only strengthened this acquiescent foreign policy stance (Lidegaard 2003). Danish politicians agreed that effectively defending the country against German military occupation was impossible. But there was widespread consensus that they should try to shore up the country against those extremists on the right who increasingly found inspiration in events south of the border, and who might conquer the country from within.

A two-pronged strategy was pursued. Firstly, broad political settlements on economic and social reforms were achieved in the years of the Great Depression, culminating in the comprehensive Kanslergade settlement of 30 January 1933 (the same day Hitler was appointed chancellor). So, in spite of the growing economic problems and a rapid rise in social grievances, no one within the political establishment attempted to take advantage of the situation; instead the four dominant parties cooperated, through pragmatic policymaking, to regain control of the economic and social situation (Kaarsted 1991).

Secondly, all signs of political extremism were effectively dealt with at an early stage. The Social Democratic head of government, Thorvald Stauning, was strongly committed to building an all-national bulwark against radical movements. Though the pressure to establish worker and youth militias grew during the early 1930s, Stauning stood firm. Likewise, the leader of the Conservative Party, John Christmas Møller, fought intensely the radicalization within his own party’s ranks. When the Conservative youth organization of the party from 1933 made Nazi-inspired uniforms obligatory for members, and fights between socialist and conservative groups became more frequent, Christmas Møller unambiguously condemned these quasi-fascist tendencies, and in 1936 he succeeded in reinstalling a more moderate leadership of the Conservative youth organization. In addition to these protective measures against specific threats to Danish democracy, parliament from 1933 to 1939 introduced legislation prohibiting uniforms in political organizations, tightening the requirements for those carrying weapons, increasing punishments for instigating social disorder, and criminalizing anti-Semitic statements (Lauridsen 2003: 75–7; Lidegaard 2003: 279–81, 331).

This commitment to the so-called ‘internal defence’ of Denmark meant that political extremism never won a bridgehead among the major players in the strongly consolidated Danish party system, that political violence remained low and that, though both a Nazi Party and a Communist Party did eventually gain representation in the Danish parliament, popular support for these, and other, anti-democratic alternatives remained marginal throughout the interwar years (Lauridsen 2003: 72–5; Lidegaard 2003: 328–30).

A Wider Comparative Perspective

Broadening our perspective to include the remaining interwar countries gives us a crude impression of whether the trends identified in the three case studies were exceptional.

In East-Central Europe in the immediate aftermath of the First World War, liberal democracy was broadly considered the superior form of government. Western democracy was pitted against a backward order of quasi-feudal authoritarianism, a spirit of democratic progress flourished and ‘Western guarantees to the new states created by the Versailles Treaty went hand in hand with the promotion of Western constitutional models’ (Rupnik 2000: 117).

From 1922 onwards, we find some evidence of Mussolini’s actions in Italy inspiring right-wing groups elsewhere in East-Central Europe and the Balkans, but mostly with a relatively modest effect on national policies (Borejsza 1980; Payne 1996: 134–45). The coups by the Military League in Bulgaria (1923), Ahmed Zog in Albania (1924), Theodores Pangalos in Greece (1925) and King Alexander in Yugoslavia (1929) were clearly inspired by Mussolini’s operational example (D’Agostino 2012: 174). But although fascist movements emerged in most of these cases, these groups – and fascist ideology – only played minor roles in the 1920s. Moreover, there was, generally speaking, not much direct pressure from the great powers to either uphold or undermine democracy. International influence consisted more in showing that a third alternative to democracy and communism was viable and that dictatorial powers could be gained by rightist forces without much struggle. However, democratic institutions were not completely discredited and even in countries where democracies had in reality broken down, there would still be multiparty elections and only moderate repression of civil liberties.

A process of radicalization ensued in the 1930s (Mann 2004) as the domestic and diplomatic successes of Nazi Germany in the 1930s created the ‘impression that authoritarian dictatorship was the future’ (Bideleux and Jeffries 1998: 468; Rothschild 1974: 21; Seton-Watson 1945: 378–85). Without much direct interference from the outside, Estonia and Latvia experienced coups in 1934, Metaxas took dictatorial powers in Greece in 1936, and Zar Boris II and King Carol II, respectively, established royal dictatorships in Bulgaria (1934) and Romania (1938).

It is worth noting that the regime types of most of the ‘little dictators’ of East-Central Europe and Portugal, though closer to that of Mussolini than that of Hitler, presented traditional versions of dictatorship that turned increasingly repressive – rather than attempts to set up genuine fascist regimes (Lee 2000; Luebbert 1991; Payne 1996). Indeed, it became a general pattern after 1933 in East-Central Europe that ‘authoritarian elites pre-empted, repelled, or expelled their Right-Radical challengers’ (Rothschild 1974: 380). To illustrate, the democratic breakdown in Estonia can be characterized as a pre-emptive coup by mainstream conservative forces that were confronted by increasingly resilient extremist forces to their right (Rothschild 1974: 372–7). It was the right-radical League of Freedom Fighters, which performed surprisingly well in elections in 1933–4, that spurred the leader of the conservative Agrarian Party, Konstantin Päts, to first seize power and, in December 1935, quell an uprising by the Freedom Fighters. This move was supported by the Estonian Social Democrats, even though it resulted in the suppression of democratic rights. In Latvia, similar dynamics operated, this time triggered by the success of the right-radical Thunder Cross. Here, all parties were banned as President Ulmanis moved strongly to the right in 1936.

Indeed, repression also increased in the 1930s in countries that had experienced democratic breakdowns in the 1920s. Take for example Lithuania, where the undemocratic President Smetona in June 1934 smashed a coup attempt by the right-radical Iron Wolf movement under the former premier Voldemaras, and Hungary, where democracy was never installed in the first place (see Lee 2000).

Turning to West-Central and Southern Europe, we identify a rather similar pattern. During the period of challenged liberal hegemony, we hardly find any signs of direct anti- or pro-democratic pressure from the outside. However, demonstration effects can be identified. In Germany, Mussolini was greatly admired by Hitler, who referred directly to him as an influential source of inspiration for strategy and ideology both shortly before the failed Beer Hall Putsch in 1923 and during the ensuing trial (Jäckel and Kuhn 1980: 729, 1027, 1198). Tellingly, one week after the March to Rome, Herman Esser, one of the founders of the Nazi Party, declared that ‘Germany’s Mussolini is called Adolf Hitler’ (Kershaw 1998: 180).

With respect to the coup in Spain in 1923, Primo de Rivera ‘called Mussolini his inspiration and teacher’, and ‘the Fascist regime always remained the nearest thing to a model for [him]’ (Payne 1998: 100–2). Nonetheless, only in 1929 did Primo de Rivera indicate that his Unión Patriótica should be reorganized along the lines of the Fascist Party in Italy, and this change was not implemented before the advent of the Second Republic. Only in the 1930s did Spain genuinely experience increased radicalization (Bernecker 2000). Monarchist parties established contacts with the fascist regime in Italy in 1934 and even received assistance to initiate a coup. Moreover, Primo de Rivera’s oldest son, José Antonio, headed a new fascist movement (Falange Española), which was supported financially by Italy. It was to become the sole fascist movement in Spain but it was relatively insignificant until General Franco made it the organizational basis of his new regime party in 1937 (Payne 1998: 104–8). Although Italy and Germany did not directly engineer the military revolt that inaugurated the Spanish Civil War, they quickly decided to intervene on the side of right-wing forces. The military assistance from Hitler and especially Mussolini, combined with the neutral position of the democratic great powers, was an important factor in Franco’s military victory in the civil war (Payne 1998: 107–9).

In Czechoslovakia we also find clear evidence of inspiration from the outside, again principally in the 1930s, but first and foremost among the non-Czech minorities in general and the Sudeten Germans in particular (Seton-Watson 1945: 171–85).

Turning finally to North-Western Europe, several of the countries that entered the interwar years with resilient democracies deepened their democracies by extending the suffrage and/or consolidating the parliamentary principle in the immediate aftermath of the First World War. When democracy was challenged from the right in the rest of Europe, not much support could be found in the north-western countries for fascist ideology or to use coups to pre-empt a left-wing revolution. The liberal past of these countries and the willingness of moderate political parties on the left and right to compromise and cooperate left little room for extremists (Luebbert 1991).

A telling example is the Netherlands, where the deep economic crisis of 1929–33 as well as the advent of Nazism in neighbouring Germany strengthened rather than weakened the democratic commitment of the main parties – and the latter development undercut support for the already small fascist party (Aarebrot 2000). Developments in Switzerland, Sweden and Norway were rather similar to the Danish and Dutch cases (Lindström 1985). Fascist currents, often inspired by the developments in Italy and Germany, were stronger in Belgium, where the Rexist movement won sizeable representation in parliament in 1936. Nonetheless, the Belgian Catholic Party together with the Liberals and/or the Socialists kept their dominant position and did not radicalize (Berg-Schlosser and De Meur 2000).

A similar pattern characterizes Finland, where we find some evidence of demonstration effects in the late 1920s and early 1930s, underwriting the explosive growth and influence of the extremist rightist Lapua Movement. However, unlike in Austria and Spain, the Finnish Social Democrats did not flirt with undemocratic behaviours and models, and the main democratic actors stood firm once the threat from the Lapua Movement became imminent. The challenge posed by the Lapua Movement did force the Finnish government to respond with some repression – first and foremost of the movement itself – but within the context of democratic survival (Alapero and Allardt 1978).

Conclusions

We have presented a theoretical framework that is based on a multifaceted view of external influences on democratic development. This framework appreciates both demonstration effects and power politics, and the anti-democratic effects of external influences are understood as conditional on the domestic vulnerability of democracy.

Our empirical analysis of cases from interwar Europe identified some anti-democratic demonstration effects in the 1920s after the takeover by Mussolini’s fascists in 1922. But we have also shown that these effects became much stronger after 1933. Furthermore, the empirical analysis showed that demonstration effects loom much larger than actual outside pressure, even in the latter period. The only partial exception is Austria from the late 1920s onwards as Mussolini attempted to bring this small neighbour country into the fascist camp and as Hitler pushed for Nazi rule and Anschluss, which was first resisted and then accepted by Mussolini for strategic reasons. Other than that, domestic polarization was nourished via examples rather than pressure, or more particularly by the surge of the extremists on the right, who referred to the danger of a revolution from the left and were clearly inspired by what happened in Italy and later in Germany.

We hold that this pattern is a consequence of the juncture of 1922 merely producing a change on the passive dimension of the international order. From 1933 both the passive and the active dimension pushed in an anti-democratic direction, furthering democratic regressions in all countries in which democracy was weak due to inauspicious domestic circumstances. Hence, through the interwar years, pro-authoritarian international influences gradually became more diverse, abundant and powerful, increasingly inspiring and in the latter part of the period even assisting anti-democratic rightist forces in gaining and consolidating power.

Shifts in the balance of power between democratic and non-democratic powers thus fed the impression that autocracy was the model to pursue. Belying Boix’s (2011) model of international influences, we have come across no evidence that Great Britain and France (or the US, for that matter) actively supported democracy in smaller countries when it came under pressure. Interwar great powers navigated in terms of narrow national interests and geopolitical rivalries rather than democracy. It was the examples set by these great powers and other states rather than hegemonic interventions that mattered. This does not mean that power politics is irrelevant, but its primary effect on regime developments seems to be through inspiration, not imposition. More precisely, a radical upheaval in the international order was needed to produce anti-democratic demonstration effects with real bite. Our empirical analysis indicates that, though we do register the demonstration effects unleashed by Mussolini’s takeover in 1922, the anti-democratic external influences only came to the fore with Hitler’s rise to power in Germany.

These empirical dynamics not only point to the need to reformulate and integrate dominant perspectives, they also have important consequences for the effects of external influences in other periods, such as the current one. Our framework and findings indicate that today’s dominant pro-democratic demonstration effects are anchored in cold geopolitical realities, namely that most great powers (and the single super power) are democratic. Were this to change, the Western liberal hegemony which has characterized the post-1989 period (Levitsky and Way 2010) would be replaced by an international order that is less conducive to democracy. Such a change could either be a consequence of democratic breakdown in established great powers or of the rise of autocratic great powers, bent on obtaining their place in the sun. The second possibility seems to be the more realistic one, with developments in China and Russia being particularly relevant. However, it also seems fair to say that major changes in the international order are not to be expected in the short term.