Case I: Group Identification Always Occurs in G

This useful benchmark occurs when v ⩾ V(sⁱ ) + sⁱ for all i = p, ν, l. As shown in (1), in this case the internal equilibrium in country G is always group identification. Then, payoffs under Peace and War for G are

For Peace to occur, we need to have both VF P ⩾ VW F and V GI P ⩾ V GI W .Footnote 8 Given risk neutrality and the allowance for transfers from one country to another the problem becomes one of VF P + VP GI ⩾ VF W + V GI W , or that the total payoffs under Peace are at least as high as those under War, which given (3)–(6) imply

Since for c > 0 this inequality is always satisfied (as a strict inequality), there is always a transfer t that would make the payoffs of both players at least as high as those under War. That is, when group identification always occurs in country G, Peace in the international conflict is always an equilibrium because victory or loss in War do not affect the domestic conflict’s payoffs and equilibrium.

Case II: National Identification in G occurs only after victory in War

This case occurs when V(s ^ν) + s ^ν > v ⩾ max{V(s^p ) + s^p , V(s^l ) + s^l }. In this case the internal equilibrium in country G is national identification after victory in War and group identification under Peace or after loss in War. Then, the payoffs under Peace and War in this case are

Again, given risk neutrality and the allowance for transfers from one country to another the problem becomes one of VF P + VP GII ⩾ VF W + V GII W , which implies

Given that in this case V(s ^ν) + s ^ν > v, the right-hand side of (8) is positive and, therefore, there are always low enough costs of War (i.e., c) for which Peace is never feasible. Note that there is no role for the disputed territories (d) in creating any incentives for War. Even if d were 0, country G would have an incentive to instigate War (provided of course that the costs are low enough) in order to reap the benefits of national identification.

There are three additional cases (examined in the Supplementary Appendix): Case III, in which Group identification occurs only after loss in War; Case IV, in which Group identification occurs only when there is Peace; and Case V, in which war always occurs. The results from all cases are summarized in Proposition 1.

Proposition 1 (i) When group identification is always the equilibrium in G (case I), Peace is always feasible.

(ii) When national identification is the equilibrium after victory in War and group identification is the equilibrium under Peace in G (cases II and IV), War always occurs when its costs (c) are low enough. The higher is the probability of victory (p) for G and the higher is the status (σ^ν) and the lower the distance cost under victory (Δ^ν), the less likely is Peace.

(iii) When group identification is the equilibrium in G only after loss in War or when national identification is always an equilibrium in G (cases III and V), Peace is not feasible only if the probability of victory (p) for G or the status minus distance cost under victory (s ^ν = σ^ν − Δ^ν) are high enough.

In sum, when we add to the standard realist setting of interstate competition under anarchy a psychological mechanism linking war, status, state capacity, and domestic intra-group conflict, we can account for war in situations where it would otherwise not occur. This mechanism applies to settings where interstate conflict occurs in the shadow of domestic conflict. Such domestic conflict is implicit whenever individuals’ allegiance is divided between nested social identities and subnational social groups compete over resources. Interstate conflict can shape patterns of domestic conflict by pushing individuals to identify nationally or parochially (along ethnic, regional, or class lines). When individuals identify with their ethnic or other narrowly defined social group, this can diminish the state’s ability to mobilize resources or public support for interstate war.

The scope conditions for the theory are broad. It captures the tradeoffs between domestic and external conflict in multiethnic states, federations, conglomerates, or empires where individuals’ allegiances do not always align with the state and where domestic conflict over resources, autonomy, or self-determination can weaken the state’s power as it competes in the arena of international politics. We mainly require that the national identity be open to any group in the country; and for nationalism to exist as an ideology (so that national identification would be meaningful to social groupings). Given that nationalism has been a motivating force of politics since at least 1815, these scope conditions are not limiting.

A basic intuition behind the model—the ebb and flow of national identification as a function of changes in national status and national power—applies to a large array of cases in both contemporary and historical settings. Four generic scenarios illustrate this variation. First is when fear of status loss leads to bellicosity, as in the examples from WWI with which we introduce this article. In Russia’s decision-making for war in 1914 we see clear evidence of our mechanism that is inconsistent with diversionary war theory. As Geyer (Reference Geyer and Little1987, 317) establishes, Russia’s bellicosity was not an attempt to divert the public’s attention from domestic failings; rather it reflected the desire to sustain the nationalist idea that was central to the elite’s strategy for generating power internationally.

Second, military defeat can generate centrifugal pressures.Footnote ⁹ The very situation leaders in Vienna and St Petersburg feared—the rise of local nationalism at the expense of identification with the larger empire—appeared with a vengeance after defeat in 1917–18. A more modern example comes from Egypt, where the resurgence of Islamism was the direct result of, among other factors, losing the 1967 war. The Six Day War undermined the promise of nationalism throughout the Islamic world: “In confronting the humiliation, the deprivation, and the utter perplexity that followed what seemed an incomprehensible military defeat, an incisive moral verdict was discerned” (Gaffney Reference Gaffney1992). Islamism gained ground at the expense of Egyptian nationalism (Ibrahim Reference Ibrahim1988); losing Arab Jerusalem “produced shock waves affecting both Arabs and non-Arab Muslims in engendering feelings of fear, insecurity and inferiority” (Dekmejian Reference Dekmejian1980, 8).

Needless to say, defeat can generate the material deprivation of both state and society, which can also feed centripetal forces. Hence, the third kind of scenario may be especially interesting: when unexpected performance against a much more powerful opponent allows national leaders to frame the loss as a victory. Here, the war’s material effects run counter to the psychological mechanisms we highlight. In the Suez War of 1956, for example, Egypt was defeated on the battlefield but lost no territory and stood up to “imperialists,” seeming to many in the Arab world to have forced the British to withdraw. The result was a major blow to Britain’s international status, a boost to Egypt’s, and an upsurge of Egyptian nationalism and support for Nasser, whom Suez made into a pan-Arab hero.

Fourth is “nation building through war,” in which military victory increases status and national identification. Material gains sometimes accompany victory in war; but extant approaches have focused almost exclusively on material sources of social identification and have assumed away the more intangible psychological dimensions that our model highlights. Though it has yet to be the subject of sustained research, abundant anecdotal evidence suggests that victory in the “Great Patriotic War” helped forge a Soviet identity after 1945 despite almost incalculable material losses. Although a Soviet “national” identity never supplanted already strong ethnic identities among many of the USSR’s composite groups, allegiance to the state and feelings of pride among Soviet citizens were high after the victory and during the early Cold War days as the USSR scored diplomatic and other victories in its competition against the rival superpower.

Similarly, China’s involvement in the Korean war helped “enhance communist control of China’s state and society and to promote China’s international prestige and influence” (Jian Reference Jian1994, preface). Consensus opinion among China scholars is that “Fighting U.S.-backed United Nations troops to a standstill in Korea added enormously to the CCP’s prestige” (Lieberthal Reference Lieberthal2004). The war “gave rise to a new mystique of Chinese endurance and heroism, which was elaborated in the People’s Republic by an outpouring of literature, films, plays, and tales of the model soldier-heroes that reinforced the values of sacrifice and revolution” (Spence Reference Spence1990, 505). Mao’s revolutionary romanticism was supported by the adversarial relationship with the United States (Zhang Reference Zhang1995, 253–4) and China’s rising prestige fed a surge in national identification among large segments of the population during a difficult period with pockets of intense conflict and facilitated the state’s propaganda machine as well as a series of coercive policies designed to unify the nation.

Although they may manifest themselves in many settings, the mechanisms we examine here are arguably clearest when a nucleus state attempts to enhance its size and power via inclusion of potential co-nationals outside its borders or when a nucleus state seeks to retain the allegiance of subjects with a potential exit option. As the fatal interaction between Serbia and the Austro-Hungarian empire in 1914 illustrates, this dynamic was strongly present in 19th- and 20th-century Europe, which frequently pitted dynamically rising national states against declining polyglot empires. Three major cases of nation-building through war stand out: Piedmont/Italy, Prussia/Germany, and Serbia/Yugoslavia. In the mid-19th century, Piedmont fought wars against Austria to expel it from Italian affairs and exploited a reservoir of common in-group identity among Italians. After losing in its first effort, it then allied with France and Britain against Russia and Turkey in the Crimean War. As Cardoza (Reference Cardoza and Davis2000) observes, although the war “produced few immediate gains for the Piedmontese, their participation in the military operations and the Paris peace conference elevated the stature of the Savoyard state on the peninsula . . . [which] triggered a surge of pro-Piedmontese sentiment among both moderates and segments of the democratic left in other Italian states that found expression in the National Society, an organization launched in 1857 to promote Savoyard [Piedmont] leadership of the independent movement.”

A half-century later, Serbia presented itself as the “south Slavic Piedmont.” As Dedijer (Reference Dedijer1966, 76) observes “despite different cultural and religious influences, there existed a feeling of unity among South Slav peasant masses.” Belgrade pursued an expansionist, nationalistic policy, enlarging its territory by 80% and population by 50% in the Balkan Wars in 1912 and 1913. Serbian leaders and intellectuals presented their nation as the potential nucleus of a much larger state that would bring together both Serbs and other south slavs, most of whom were Austro-Hungarian subjects. Authorities in Vienna attempted to counter by stoking particularistic and local identities in the lands they administered but, as noted, Serb battlefield success in the Balkan Wars “pitched Serbophilia to unprecedented euphoric heights” among many Croats in the Austro-Hungarian Empire (Banac Reference Banac1984, 103). In part owing to its vigorous participation as an active belligerent in World War I, Serbia emerged as a dominant partner in the Kingdom of Serbs, Croats, and Slovenes formed in 1918.

Though each case is complex and the relative significance of war to state-building varies, each featured a nucleus state that could plausibly appeal to a larger national identity. In each, elites in the nucleus state were motivated to expand in part by realpolitik concerns of power and security. In each, the nucleus state engaged in war not just to expand territory but to gain status and thereby enhance the appeal of the larger national identity. In each, evidence suggests that military victory affected social identification among key populations and facilitated the construction of larger state institutions. This strategy would find its most famous expression in the person of Otto von Bismarck and the case of German unification.

If Germany were to Remain Fragmented

To show how the prospect of German unification changes arming, we first consider the benchmark counterfactual case of a fragmented Germany regardless of the war outcome. In the second War-or-Peace stage in which arming has already been determined and is “sunk” the payoffs under War (w) and Peace (p) are as follows:

where τ denotes the size of the transfer from F to G. In this case the total surplus under War (V^w _F + V_P ^w ) equals d + v − 2c, which is strictly lower, because of positive War costs, than the total surplus under Peace (V^p _F + V_P ^p = d + v). Therefore, in the presence of transfers, there is always a way to have Peace. Arming, however, will still take place in order for each side to maintain a better bargaining position in the presence of transfers. The transfer itself is determined by the bargaining process, which we suppose involves an “equal splitting of the surplus” between the two countries so thatFootnote ¹⁴

Note that the higher Prussia’s share of the disputed territory (β), the lower is the implied transfer from France since Prussia is already sufficiently compensated from already holding the territory. Moreover, the implied transfer is higher the higher Prussia’s winning probability is. Since the transfer in the event of Peace depends on the probabilities of winning and the probabilities of winning depend on arming, it becomes evident why the transfer is a function of arming and why arming can be expected even when Peace is always expected.

By substituting the transfer τ* into the Peace payoff functions, we eventually obtain what the two countries face in the first stage in which arming is chosen:

The Nash equilibrium choices of arming are identical for the two countries in this case:Footnote ¹⁵

The level of arming depends solely on the value of the disputed territory and none of the other variables. The two countries have equal probability of winning in War but peace is always an equilibrium.Footnote ¹⁶

When Germany can Unify

Next, we return to the case in which victory after War would bring about German unification, whereas it would not do so under either Peace or after loss in a War. As discussed in the previous subsection, the payoff for Prussia after victory is now V(s ^ν) + s ^ν + d/2, with the payoffs under the two other states remaining the same.

Then, in the second stage of the game, after arming has already taken place and the choice is between War and Peace, the payoffs for the two countries are

While France’s payoff functions remain the same, the payoff of Prussia under War receives the additional boost of p[V(s ^ν) + s ^ν − v − d/2] which reflects the extra economic benefit of German unification and the higher status and lower regional frictions that would follow a victory in War. Now the total surplus under War ( $\overline{V}_{F}^{w}+\overline{V }_{P}^{w})$ equals d + v + p[V(s ^ν) + s ^ν − d/2 − v] − 2c and is no longer guaranteed to be higher than the total surplus under Peace ( $\overline{V} _{F}^{p}+\overline{V}_{P}^{p}=d+v).$ In particular, if

(that is, if the expected extra benefits of War to Prussia are greater than the total war costs of the two countries), then no transfer from France to Prussia could prevent War in the second stage of War and Peace.

Status, Social Identity, and the Effects of Intra-German War

In the years before 1870, evidence available to Bismarck and other top Prussian leaders indicated that unification would not occur without the nationalist upsurge that a victory over France would produce. To enhance its power Prussia had three strategies: conquer other German states by force; use subtler incentives to draw the southern Germans into the fold; or increase the attractiveness to Germans of identifying with a German Reich that did not compromise Prussian identity by enhancing Germany’s interstate status. Ever the pragmatist, Bismarck tried all three. After defeating Austria, Prussia conquered and annexed several northern German states in 1866, and then used war booty to enhance material incentives for unity. Ethnic conflict makes ethnic identities more salient and fighting German-led Austria and its German allies (Bavaria, Württemberg, Saxony, Hanover, Baden, and several smaller German states) generated a backlash of hostility toward Prussia and a major setback for the cause of unification. Bismarck then tried using economic inducements via the customs union, but southern German particularists used the opportunity to increase their sway, producing major defeats for pro-unification parties at the polls (Pflantze Reference Pflantze1990, 392–5). After Southern voters elected antiunionist conservative candidates in the Customs Union elections of 1868, a Swabian newspaper proclaimed the vote as “the protest of South Germany against Prussification.” Bismarck himself took the result as a clear signal “that the south wishes to have no further connection with the north beyond customs treaties and the alliance treaties.”Footnote ¹⁸ Off the record he agreed with his envoy to Munich that “German unity is likely to be fostered by violent events,” but stressed that “it is self-evident. . .that Germany at this moment is not a ripe fruit. . . .” Footnote ¹⁹

The strategies of conquest and subtle inducement failed. As Carr (Reference Carr1991, 172) writes, “on the eve of the Franco-Prussian war anti-Prussian sentiment in the south was. . .growing stronger, not weaker.” The local parliamentary elections of 1869 ended up with the sound victory of particularist parties. In Württemberg, particularists won a majority and, encouraged by a petition endorsed by 150,000 signatories, pressured the government in freeing Swabia from the “slavery of the barracks” by cutting military expenditures and into reducing the length of military service, against Prussia’s wishes. In Bavaria, the Patriotic Party won a majority and passed a no-confidence vote on the Prussian-leaning prime minister Prince Hohenlohe to replace him with a pro-Austrian cabinet (Pflantze Reference Pflantze1990, 405–8).

In sum, Prussia’s status gains at Austria’s expense were not enough to offset the widening social distance that this intra-German war created. There was thus a weakening of national (German) identity, while the expected payoffs to unification persisted and continued to motivate Prussian elites to pursue that goal.

Status, Social Identity, and War with France

Bismarck’s deteriorating position on the intra-German chessboard led him to pursue an aggressive strategy to gain prestige at the expense of France, even at risk of war. As he put it to King Wilhelm, “every recognizable attempt of Prussia to determine the decision of the south-German princes, by pressure or agitation will have as its consequence the opposite of the result sought” and would have strengthened rather than weakened the ties between Bavaria and Wurttemberg and Paris and Vienna (Böhme Reference Böhme1971, 203–4; Wetzel Reference Wetzel2001, 63–71). The strategy of inducement necessarily involved gaining prestige at France’s expense, which Bismarck ultimately sought to accomplish by encouraging the candidacy of a Prussian Hohenzollern prince for the newly available throne of Spain (Steefel Reference Steefel1962, 56–88; Steinberg 2011, Chap. 5; Wetzel Reference Wetzel2001, 77–88). As the south German states’ major remaining great power patron, France was an obstacle to unification and, by presenting it with a setback, Prussia would undermine the confidence of Munich and Stuttgart in French commitments to their independence while raising the prestige of Berlin among southern Germans.

Bismarck set forth his reasoning in a letter to King Wilhelm in March 1870: “it is desirable for Germany to have on the far side of France a country on whose sympathy she can rely and with whose susceptibilities France would be obliged to reckon. . . . The repute of the Hohenzollern dynasty, the justifiable pride with which not only Prussia regards its Royal House but Germany too—she tends more and more to glory in that name as a common national possession, a symbol of German fame and German prestige abroad; all this forms an important element in political self-confidence, the fostering and strengthening of which would be of benefit to nationalist feeling in general” (Böhme Reference Böhme1971, 218–9). The evidence suggests that Bismarck’s preferences were shared widely within Prussia’s ruling elite.

Consistent with Proposition 2(i), the high expected benefits of German unification led Prussia to arm proportionally more than France. Chief of Staff Moltke and War Minister Roon had been assiduously organizing the army to prepare for such a war since 1866 (Howard Reference Howard1962; Wawro Reference Wawro2003). In each preceding crisis Bismarck favored negotiated settlements with France, brushing aside Moltke’s concern that France might finally reform its military establishment and present a tougher challenge. He only came to the tough line against France when other routes to unity seemed to be running into a dead end.

Bismarck’s plan to present Paris with a fait accompli was compromised when the Hohenzollern candidacy became public information, sparking an immediate French demand for its revocation. In the ensuing complex diplomacy numerous compromises were negotiated that would have saved both French and German face. Bismarck worked feverishly to scuttle such deals, making sure that the crisis was framed as a status contest between France and Prussia. “Bismarck. . . set sail on a collision course with the intent of provoking either war or a French diplomatic humiliation” (Pflantze Reference Pflantze1990, 462).

In Paris, Emperor Napoleon III and his foreign minister Antoine de Gramont also faced a choice. Key here is that they had the same preference ordering as prevailed in Berlin: better war than a mutually face-saving deal that preserved the status quo. The French were determined “to turn the affair into a public humiliation of Prussia” (Price Reference Price2001, 427). A key desideratum was to deal a blow to Prussia’s rise by thwarting its further expansion at the expense of the southern German states (Echard Reference Echard1983; Price Reference Price2001; Wawro Reference Wawro2003, 9–10). In the view of leading French decision makers—shared in other European capitals—France’s prestige relative to Prussia had declined after many diplomatic and military setbacks, notably the smashing of France’s Austrian ally and then Bismarck’s brusque and ultimately public rejection of Napoleon’s attempt to gain compensation by acquiring Luxembourg (Case Reference Case1954, Chap. 9). The French emperor’s advisor warned that “Grandeur is relative” (Wawro Reference Wawro2003, 17) and France was not in a position to risk further decline relative to Prussia. Eugene Rouher, French minister of state, had a clear strategy in mind: “Smash Prussia and take the Rhine. . . . By ‘the Rhine,’ Rouher meant Prussia’s western cities: Cologne, Dusseldorf, and the Westphalian Ruhrgebiet around Essen, Dortmund, and Bochum” (Wawro Reference Wawro2003, 17). Even the prime minister, Ollivier, a socialist originally opposed to Napoleon’s imperialism, ultimately rejected a face-saving solution as unbefitting France’s “honor” (Wetzel Reference Wetzel2001, 145). As the crisis gathered steam, evidence poured into the government of public support for a hard line against Prussia (Case Reference Case1954, Chap. 10). Thus, our proposition 2(ii) illustrates the motives for escalation to war both within Germany and France.

Germany’s domestic conflict played into French calculations. Case (Reference Case1954) shows evidence of concern in Paris that backing down to Prussia would likely pull the rug out from under pro-France and anti-Prussian southern Germans. Should France defeat Prussia militarily, Prussian expansion in Germany would not only be halted but probably reversed as recently annexed states like Hanover would seek greater autonomy. To be sure, should Prussia win, a consolidated Germany was virtually certain, but decision makers in Paris still gave France a clear edge (Carr Reference Carr1991; Howard Reference Howard1962; Wawro Reference Wawro2003). Pre-war French assessments projected a groundswell of support in non-Prussian Germany (Price Reference Price2001, 440). As the Grand Duke of Hesse assured one French general, “the slightest [French] success will determine all the states of South Germany to march with you” (Howard Reference Howard1962, 45).

The War’s Effects on Social Identification in Germany

War with France generated two main psychological effects for German unity: a rally-around-the-flag effect against a common external foe, and the increased salience of the new Prussian-sponsored German identity in light of its heightened interstate prestige. The rally-around-the-flag effect is notable before and in the early stages of the war, when the war's outcome was unknown, and it facilitated large-scale mobilization and arming that gave Prussia an edge it would otherwise not have had. The status effect was crucial for unification and becomes clearer as the immediate threat from France recedes and Germany’s augmented international status becomes evident after a series of decisive victories.

Evidence abounds of widespread popular support for Prussia against France, such that the governments concerned had essentially no choice but to honor their treaty obligations to fight with Prussia or risk popular revolt (Böhme Reference Böhme1971, 231–2; Wetzel Reference Wetzel2001, 155–8). There were still hold-outs, however, especially in Catholic-heavy areas. Thus, in Bavaria peasants cut their corn green so it would not be trampled by Prussian troops, and expressions of sympathy toward France could be heard from parishes to the parliament, where many members of the winning party actually favored neutrality (the vote on the military budget which was the best indication of support for Prussia won by 100 to 47 votes). Support for France could also be detected in recently Prussian-occupied areas such as Hanover and Mainz, where the population was “waiting for the French as if for the Redeemer” (Howard Reference Howard1962, 59–60). But, on the whole, the Southern states’ populations supported the war due to genuine nationalist sentiment. The Bavarian People Party thus had to bow to popular demands and demonstrations of nationalist fervor occurred in both Bavaria and Wurttemberg (Fink Reference Fink, Speirs and Breuilly2005, 155–71; Howard Reference Howard1962, 59–60; Pflantze Reference Pflantze1990, 491). According to one historian, “Though the princes, soldiers, and bureaucrats of south Germany had a vested interest in remaining outside of Prussia, millions of south German citizens wanted a nation-state, which was something only Prussia could deliver” (Wawro Reference Wawro2003, 24).

The effects of military success can be seen more clearly in the decision to create the German Reich. The decision was taken in the aftermath of the crushing victory at Sedan, which determined the outcome of the war. The minister president of the North German Confederation Rudolf von Delbrück reports on a conversation on September 7 with King Wilhelm and Bismarck that the creation of a united Germany was now seen as an “easy task” after eight victorious battles in which the German forces had fought together. As Delbrück reports, the feeling was that “self-confidence, heightened by great deeds and great success, outweighed the army’s feeling for the individual states whose contingents composed it. This self-confidence overshadows or overcomes every other feeling. He who did not consider feeling for Bavaria or for Württemberg as impermissible now and would not hear of ignoring or overcoming state particularism, ran the danger of passing for a dull or mean man, if not for something worse” (Böhme Reference Böhme1971, 237–78). This was an argument that would have been anathema for Bavarians and Swabians at the start of the year 1870, and that Bismarck could not have voiced even at the beginning of the hostilities with France. To the extent that it could be stated outright, it was only due to the changed circumstances created by military success.

Bismarck eschewed any hint of coercion versus the southern German states, which now lacked any great power protector.Footnote ²⁰ There was no reason not to offer generous terms of autonomy, since, as he put it, “the Rubicon has anyhow been crossed.” To preserve the Prussian identity he treasured, as well as to smooth the path for southern German entry into the new Reich, Bismarck did not seek a monolithic national state similar to France; instead, as he put it, he preferred to “absorb. . .German nationalities without nullifying them” (Schmidt 1985, 41).

What we observe is increased salience of the German identity even as individuals retained their local identities (see, e.g., Confino Reference Confino1997; Applegate Reference Applegate1990). As Green (Reference Green, Speirs and Breuilly2005, 132) argues, the German nation was understood as “a composite nation, made up of several sub-national groups.” Individuals retained some loyalty to different identities each nested within a more encompassing one (cf. Wimmer Reference Wimmer2008). But in the period that we are analyzing, German identity assumed a more concrete form, defined by political developments on the ground, just as constructivist theorists would predict. Thus, the observed shift in social identification was toward the Prussia-dominated German Reich Bismarck had been constructing ever since the victory over Denmark in 1864. The two identities were no longer seen as contradictory. To quote Hardtwig (Weichlein Reference Weichlein and Retallack2000, 167): “the rivalry between the nation and states diminished in significance. It seemed to be more important that national consciousness came to be based on particularistic identities; the supportive aspect overshadowed the rivalry.” This cognitive shift is consistent with the re-alignment of social identification due to war in the model. As strong as local identities remained in Baden, Saxony, Württemberg, and Bavaria, the victory over France occasioned a major shift from fighting Prussia in 1866 to voting in anti-Prussian majorities in 1870 to celebrating entry into a Prussia-dominated Germany in 1871.

Prussia did not even have the chance to begin in earnest discussions for unification before the southern states approached Berlin of their own initiative. Before Delbrück had a chance to call in session the Zollparlament, Bavaria preempted the action by suggesting a “constitutional” or even “diplomatic” alliance between the North German Confederation and the southern states. This concession represented precisely the outcome that had been rejected tooth and nail in the 1868 vote. The victory at Sedan and the nationalist effervescence that it generated sealed the Bavarian decision. And the newfound power of the national identity supported a wave of state capacity building, as the North German Consitution was extended over the new German Reich and a vast array of new state institutions was established, from the central bank and national currency to social welfare and education. Most notable was the unification of German laws, which the governments of Bavaria, Wurtemberg, and other states initially resisted. But, Eyck (Reference Eyck1950, 197) observes, “the national movement swept these obstacles away” and paved the way for investments in state capacity.