Open economy politics and international security dynamics: Explaining international cooperation in financial crises

Introduction

Why do states cooperate during international financial crises? Given the extensive cooperation between national authorities since 2007, this is a highly pertinent question for scholars of International Political Economy (IPE) and International Relations (IR). Rather different answers are given by scholars who focus on two of the most important aspects of recent crisis management: the US Federal Reserve’s (the Fed’s) provision — backed by the US Treasury — of hundreds of billions of dollars of liquidity to 14 central banks and many private global banks; and the extensive and complex series of international rescues that have taken place in Europe. ¹ A common explanation of the former is that US authorities were motivated largely by domestic political-economy considerations, especially the degree to which domestic banks were exposed to distressed foreign jurisdictions (Aizenman and Pasricha, 2010; Broz, 2014; McDowell, 2012). In contrast, the literature on recent European rescue programs has focused much more on the role of external factors, including a plethora of international institutions and the diffusion of ideas (Braun, 2013; Glencross, 2014; Henning, 2015; Hodson, 2014; Mabbett and Schelkle, 2014; Schwarzer, 2015).

This explanatory divergence may derive from different ontological predilections among scholars focusing on policy outcomes in the US and Europe (Farrell and Finnemore, 2009) and (relatedly) the supranational nature of key European institutions such as the European Central Bank. However, there are reasons to think that different research methodologies may also be driving the divergence, with Americanists tending explicitly or implicitly to rely on an Open Economy Politics (OEP) approach and Europeanists on a more eclectic and open-ended methodology.

The OEP research program, dominant in US IPE since the mid-1990s, focuses on domestic factors in explaining foreign economic policies and reflects the general turn away from international systemic approaches, notably, hegemonic stability theory (HST). ² According to Oatley (2011), the “reductionist gamble” of the OEP program sets aside international factors in order to generate greater theoretical rigor and scientific knowledge. He shows that this reductionism sometimes comes at the cost of biased inferences.

There are theoretical and empirical reasons why OEP approaches to international financial crises might also suffer from this defect. First, financial crises can be generated and promulgated by networks of complex international interdependence with endogenous tendencies to instability and contagion (Bruno and Shin, 2014; Reinhart and Rogoff, 2009: 141–161). Second, policy responses to such crises may also be shaped by systemic factors, including the presence and nature of international institutions relevant to financial crisis management, diplomatic relations between states, and the international diffusion of policy ideas. Scholarship has shown that this has been so in the past, including for policy responses by the United States (Copelovitch, 2010; Eichengreen, 2015; Eichengreen and Temin, 2000; Thacker, 1999).

As Oatley (2011: 313) notes: “OEP scholars have paid too little attention to the frequency with which and the conditions under which domestic politics can be studied in isolation.” It is essential for theory, methodology, and empirical research in our field that we clearly specify and test such scope conditions. One plausible condition under which OEP approaches will be problematic, as recent European rescues suggest, is in highly integrated and deeply institutionalized international systems. A second is when levels of great power rivalry are rising and international insecurity is increasing. In both cases, I argue, systemic factors are more likely to shape international financial policy decisions than the OEP approach allows.

Since the high levels of interdependence and institutionalization in Europe are arguably sui generis, I set the first condition aside and focus instead on the second. Although this second condition is more likely than the first to pertain generally in international relations, it receives strikingly little attention in the literature. It is also highly pertinent in the early 21st century, when the international system is again becoming increasingly fluid and rivalrous.

My main argument in the rest of this article is that when levels of interstate rivalry and international insecurity are increasing, international financial policy is likely to be shaped increasingly by national security considerations. To test this proposition, I focus on an international system in which the level of international institutionalization is comparatively low (thereby abstracting as much as possible from the first condition): the evolving European states-system in the four decades before the First World War. ³ This was also, of course, the era of the classical international gold standard. The primary form of intra-crisis international cooperation in this system was the provision of emergency short-term loans to other European authorities, or what are usually termed “international lender of last resort” (ILLR) facilities. ILLR support during financial crises is, if anything, even more important today, as demonstrated by the US Fed’s emergency swaps and in “Troika” loans to Greece and other distressed sovereigns in Europe. Focusing on the pre-1914 European system is also appropriate given my broader concern with disciplinary methodology as the consensus view in the IPE literature on this period is closely aligned with the OEP approach: domestic factors are generally seen as dominating decisions to provide ILLR support to foreign authorities during periodic crises under the international gold standard (e.g. Broz, 1997; Helleiner, 2014: 176–177).

In contrast to this consensus view and consistent with my second condition, I show that as the European states-system evolved towards greater rivalry and insecurity in the years leading up to 1914, there was a sharp increase in the importance of diplomatic and geostrategic considerations in emergency lending during financial crises. In the 1890 Barings crisis, domestic political-economy motivations dominated for the major lender, France; the OEP framework thus generates reasonably accurate inferences in this case. However, this became less and less true over time. In a series of financial crises after 1900, rising great power rivalry and the looming shadow of war increasingly shaped international emergency lending, which was increasingly aimed at building and cementing alliances. After 1907, as insecurity in Europe worsened, most major governments prioritized the hoarding of gold as a war chest.

The remainder of this article is organized as follows. The next section briefly outlines the disciplinary turn towards the OEP approach, as reflected specifically in the IPE debate over ILLR support during financial crises, before specifying the conditions under which this approach is likely to fail. The second section outlines three phases of international financial policy under the gold standard: lending for domestic interests until 1900; lending to build and cement alliances until 1907; and gold hoarding thereafter. The final section examines the broader empirical and theoretical implications of my argument and findings.

The OEP approach and ILLR actions in financial crises

After the Great Depression, during which many thousands of banks were allowed to fail in the US and elsewhere, it became a domestic economic policy norm that central banks should provide emergency loans in financial crises to domestic banks when market finance becomes unavailable. The “lender of last resort” was seen as a solution to the problem that during systemic crises, many banks will be experiencing runs as their short-term creditors (including depositors) attempt to reduce their own exposures, so that only central banks will be in a position to inject liquidity into the banking system to prevent its collapse. The principle that central banks should always lend to solvent banks experiencing liquidity crises was popularized by Bagehot (1873), who argued that the Bank of England had accepted such responsibilities in some 19th-century financial crises but not in the deep Overend, Gurney crisis of 1866.

Crucially for the early development of IPE literature, Kindleberger (1973) extended this argument to the international level, arguing that London also provided stabilizing emergency loans to foreign counterparts during crises under the pre-1914 international gold standard (Kindleberger, 1973: 290–291). HST was, in large part, formulated as a generalization of Kindleberger’s claim that only hegemonic powers possessed both the incentive and the resources to perform ILLR and other system-stabilizing functions (Eichengreen, 1989; Keohane, 1984; Snidal, 1985). HST has been subjected to extensive critique, which need not be rehearsed here, but the decisive criticism was that it abstracted from domestic politics, especially, though not only, within the hegemonic power (Haggard and Simmons, 1987: 501–502).

This prompted a pivot towards theories of foreign economic policy based on domestic variables, in the first instance, domestic societal interest cleavages (e.g. Frieden, 1991; Hiscox, 2014). This was a key point of departure for OEP theory, proceeding from the assumption that the policy choices made by domestic authorities will primarily reflect the preferences of the dominant domestic political coalition. As applied to ILLR policy, if, and only if, this domestic coalition in creditor countries favors the maintenance of the international monetary regime will national authorities provide potentially costly loans to foreign authorities in crises (Broz, 1997: 56).

Many argued that this interest-centric approach ignored domestic institutions and ideas. For example, in the context of foreign financial policymaking, restricted electoral franchises and independent central banks could provide political insulation for domestic monetary authorities and allow internationally oriented economic elites to cooperate during financial crises (Eichengreen, 1992: 6–9; Simmons, 1994). Gallarotti (1995) argued that policy norms regarding the role of central banks in economic stabilization could also be important. For example, if central banks are accorded wide discretion in choosing stabilization policies for reasons of perceived technical expertise, they may more easily be able to provide ILLR assistance to foreign authorities during crises. Within mainstream IPE, most attention has been given to the way in which domestic institutions aggregate, shape, or block interest group preferences, thereby reinforcing the primacy of OEP approaches.

The dominance of such approaches is particularly stark in the literature on ILLR policies. One indication of this are the accounts provided in standard IPE texts. In one of these (Broz et al., 2009), Part IV on Money and Finance begins with excerpts from Broz (1997) and Eichengreen (1989). Both authors provide extensive criticism of HST and the first uses domestic societal interests theory to explain ILLR actions under the pre-1914 gold standard. Alternative approaches are scarce. Gallarotti, in a more constructivist tradition, argues that under the gold standard, institutions were embedded within, and underpinned by, a normative “superstructure” of classical liberalism emphasizing openness, stable money, and fiscal restraint that depoliticized and insulated monetary policymaking from populist politics (Gallarotti, 1995: 211–213). However, his empirical account of ILLR activity under the gold standard is close to Broz’s domestic interests theory (Gallarotti, 1995: 224, 227; see also Gallarotti, 2005: 629). The dominance of this approach in the analysis of the classical gold standard is also indicated by the similar account provided by a scholar not generally associated with the OEP research program (Helleiner, 2014: 176–177).

The popularity of the OEP approach owes much to its ability to account for the sharp empirical divergence of ILLR practice in the classical gold standard from the predictions of HST. As many authors have pointed out (e.g. Eichengreen, 1989; Gallarotti, 1995, 2005), including even Kindleberger (1984) himself, it was the French monetary authorities and, to a lesser extent, those of Germany and Russia that provided emergency loans to the Bank of England in the most severe crises of the period, not the reverse. Broz argues that French and German emergency lending to Britain was an unintentional and fortuitous by-product of a dominant private sector preference for policies that dampened the domestic business cycle and avoided interest rate volatility: “Germany came to share with France the role of international lender of last resort, smoothing out shocks to the payments systems that threatened the independence of domestic macroeconomic policymaking” (Broz, 1997: 82). This preference differed sharply from the dominant societal preference in Britain, which was to maintain the internal and external gold value of sterling even at the cost of greater macroeconomic volatility. The Bank of England’s more active use of its discount rate reflected its much smaller gold reserve compared to those of France, Germany, and Russia, and the preferences of its own dominant creditor and pro-trade coalition (see also Feldman, 1997: 29; Flandreau, 1997: 761).

In spite of this apparently strong disciplinary consensus, there are at least two important sets of circumstances in which OEP approaches to explaining international cooperation during financial crises could be misleading. The first is in international systems that exhibit relatively high degrees of integration and institutionalization. In such systems, economic interdependence and the pooling of policy sovereignty in shared institutions will shape the incentives for national authorities to engage in international cooperation during financial crises. It is more likely that regional systems will exhibit higher levels of both interdependence and policy pooling, with the European Union (EU) an exemplar and possibly unique case. As the EU has shown since 2008, the relationship between financial crises, policy responses, and international institutions can be dynamic, prompting further policy pooling and institutional innovation. However, there is no reason to assume that this process will be a linear one.

The second set of circumstances in which I expect OEP approaches to be less appropriate is when the international system is increasingly fluid, rivalrous, and insecure. There is little direct exploration of the role of security factors in shaping ILLR assistance in the literature, including in HST. There are arguments associated with realist theories that security factors can strongly shape foreign economic policies, particularly trade policy (e.g. Gowa, 1995; Rowe, 1999). More directly relevant is Kirshner’s (1995) argument that states often use currency policy as tools to achieve security objectives. When national monetary authorities are losing reserves in a financial crisis, this can create an acute form of dependence that provides an opportunity for other states with access to or control over monetary assets to influence the policies of the crisis-hit state (cf. Hirschman, 1945; Kirshner, 1995: 12–17).

The potential for linkage between financial policy during crises and state security objectives should not be surprising. Jervis, for example, insists that IPE should take more consistently into account that “international politics has always taken place in the shadow of war” (Jervis, 1998: 991). It is difficult to imagine any national monetary authority providing ILLR assistance directly to an identified enemy state. More interesting is the likelihood that the importance of security factors in financial policy in general will wax and wane dynamically, depending on context. As Jervis (1998: 991) notes, their importance diminished substantially among the community of advanced Western democracies after 1945.

My claim is that in international systems in which rivalry is increasing and states perceive that they are becoming less secure, foreign financial policy is more likely to be shaped by national security considerations. Authors who have noted that international financial regimes are embedded within an international security structure have not tended to explore how shifts in this structure could alter motivations for ILLR assistance during major crises (e.g. Gallarotti, 1995: 187–188). The source of increasing rivalry and insecurity could include rapid shifts in the balance of power or regime change in a major power. As uncertainty and insecurity increases, creditor states will be more likely to lend to allies or potential allies than to others. They might also judge that lending to potential enemies will help to deter future hostile behavior or wean such states from rival alliances. Thacker (1999) makes a related argument that highlights the importance of such dynamics in International Monetary Fund (IMF) lending: countries that are moving towards alignment with the US are more likely to receive IMF loans. In sum, as interstate rivalry increases in an international system, the risk that the OEP approach will have diminishing payoffs increases.

International lending during crises under the pre-1914 gold standard

My empirical strategy is to set aside the first condition in which international factors are likely to be more important than the OEP approach assumes and to focus on the second. This choice is driven by space considerations and by more substantive reasons. First, highly integrated and institutionalized international systems are relatively rare in international relations. As already noted, the EU may be exceptional, whereas international systems in which rivalry and insecurity are increasing are relatively commonplace. Second, focusing on the pre-1914 period, when the level of international institutionalization relevant to the management of international financial crises was much lower than after 1945, allows us to investigate the impact of security factors in comparative isolation. The international economic system is far more institutionalized today compared with the pre-1914 period, and above all in Europe. ⁴ Thus, the contemporary period exhibits both increasing insecurity and much higher international institutionalization than the pre-1914 system, making it difficult to determine the effects of both factors on foreign economic policy. Third, as we have seen, the OEP approach dominates IPE approaches to the classical international gold standard, so the latter provides a strong test of its assumptions.

The IPE literature on the pre-1914 system agrees that it was weakly institutionalized and generally portrays it as a single era of British-led international liberalism and of the first phase of globalization (Eichengreen, 2008: 32–37; Frieden, 2006: 13–27; Held et al., 1999: 192–198; O’Brien and Williams, 2007: ch. 2). The classical era of the international gold standard in particular, usually dated 1873–1914, ⁵ is seen as founded on a persistent compatibility of economic interests among European great powers. The key manifestation of cooperation during crises was emergency lending by major continental European central banks to the Bank of England (BoE), which operated with a notoriously “thin film” of gold reserves (Broz, 1997; Eichengreen, 1992: 49–54; Eichengreen and Flandreau, 1997: 22; Gallarotti, 1995: 211–227; 2005).

Other important episodes of emergency lending are largely ignored in this literature, however. One of these was the massive 1906 loan to Russia organized by the French, which enabled Russia to remain on the gold standard and which the Tsar’s key minister characterized as “the loan that saved Russia” (Yarmolinsky, 1921: 285–315). A series of smaller financial disruptions after 1907, culminating in the collapse of the international gold standard at the outset of war in August 1914, have also received little attention. I consider the nature of international financial cooperation in these episodes along with that in the more familiar 1890 and 1906–1907 crises.

As regards the European states-system, the European Concert of great powers eroded over the four decades after the Franco-Prussian war. Concerned at the rise of Germany, which had agreed a Triple Alliance with Austria-Hungary and Italy in 1882, France and Russia signed a Dual Alliance in 1894. The possibility of another war between France and Germany was on the minds of European governments at least from the first Moroccan crisis of 1905, and probably earlier. At the peak of this political crisis in June 1905, the French government cancelled all military leave and Germany threatened to sign another defensive alliance with the Ottoman Empire. The Algeciras conference of January–April 1906 found Germany largely isolated and unable to achieve its goal of driving France and Britain apart (the latter two had signed the Entente Cordiale in April 1904, which settled a number of imperial disagreements) (Andrew, 1968). In August 1907, Britain and Russia also resolved some of their imperial conflicts in a convention, paving the way for the Triple Entente between Britain, France, and Russia in the same month. After the second Morocco crisis of 1911, the European Concert lay in tatters as rivalry continued to increase. Military expenditure by the major European powers also rose steadily, more than doubling in real terms between the 1890 Barings crisis and 1913 (Eloranta, 2007: 261). I show how this deteriorating security environment increasingly shaped international financial policy in crises after 1900.

In short, my main argument is that international emergency lending under the classical gold standard cannot be seen, as in OEP approaches, as a single period driven by a common set of (domestic-level) factors. Instead, it falls roughly into three main periods: lending primarily for domestic interests (1873 to the end of the century); lending to build and cement alliances (1900 to 1907); and gold hoarding for national security purposes (1908–1914). ⁶ This dating is intended as approximate since the timing of international financial crises does not always coincide with the boundaries of these three periods. Nor should this periodization be taken to mean that lending for domestic interests disappeared as a factor in the subsequent two periods, that lending to cement alliances was not important after 1907, or that gold hoarding did not occur before 1907. However, as I show, the primary motivation of the international financial policies of the major European powers shifted in each of these three phases.

Lending to build and cement alliances: Crises over 1906–1907

After 1898, when Delcassé became foreign minister, France actively tried to bring Britain into an anti-German alliance, which required Paris to overcome persisting Russian concerns. This emerging strategy influenced France’s international financial policy, which was increasingly aimed at building and cementing its alliances.

The first major financial crisis of this new period took place over 1906–1907 and was much deeper than that in 1890. It also coincided with the fracturing of the European Concert into two rival blocs in the aftermath of the first Moroccan crisis. The financial crisis reflected the growing global importance of the large but unstable US economy. Large insurance payouts from London to the US following the San Francisco earthquake of April 1906 produced a drain on Britain’s gold reserves. The BoE raised its discount rate to 6% by October and central banks in France and Germany responded by raising rates. A US boom and in Wall Street stocks drained more liquidity from London. As this boom turned to panic in March 1907, new fears grew about the exposure of British financial houses to the turmoil on Wall Street (Bruner and Carr, 2007). The BoF had offered financial assistance of £3 million in gold in 1906, but the BoE, either out of pride or because it thought it unnecessary, repeatedly refused a French loan and instead asked the BoF to purchase sterling bills of exchange (Eichengreen, 1992: 50). The BoF discounted more than 65 million francs (£2.6 million) of sterling bills between November 1906 and March 1907, refrained from further increases in its own discount rate, and permitted gold outflows to London. The bills were reimbursed, as agreed, in the second quarter of 1907.

The autumn of 1907 saw the outbreak of renewed financial panic in the US, which spread rapidly to London. This time, the BoF and the German Reichsbank allowed their gold reserves to decline and flow to Britain, which was still losing gold to America. This outflow of gold came at some cost to the Reichsbank, which, in turn, obtained help from the Bank of Austria (Flandreau, 1997: 758). At the height of the crisis, in November 1907, the BoF purchased up to 80 million francs (£3.2 million) in sterling bills and forwarded 80 million francs in gold coin to London (Eichengreen, 1992: 52, 54).

The BoF’s justifications of its financial assistance to Britain were again pragmatic. Discussing the 1906 discounting exercise, BoF Governor Pallain explained that:

we only gave our gold where necessary, and with the certainty that it would be directed where its action would be effective, and where we had real interest, from the point of view of French trade, in preventing a possible crisis. ¹²

On 7 November 1907, when the Conseil de Régence was asked to ratify the emergency discounting of up to 80 million francs worth of British bills agreed the previous week, the minutes recorded that:

[t]he operation, surrounded by the same guarantees as that carried out last year, and similarly yielding a very comfortable benefits margin, aimed at countering the effects of the London monetary crisis and at reducing, to the greatest extent possible, its consequences for our financial markets. ¹³

As ever, the BoF felt compelled to justify its actions to its shareholders in financial terms. The Conseil de Régence, the BoF’s governing body, was under the authority of a governor appointed by the state and seconded by two state-appointed deputy governors. Three treasurers brought the total number of civil servants sitting on the Conseil to six, all of whom became shareholders on appointment. The other nine regents were all representatives of the remaining shareholders, who were the most powerful bankers and industrialists in the country and known as the “200 families.” Thus, the majority of the Conseil represented commercial interests.

Despite this, there is evidence that the changing European security environment substantially reinforced the incentive for the BoF to provide emergency loans to Britain when financial crisis threatened London (Plessis, 1998: 135; 2000/2001; Vignat, 2003: 338). The French Ambassador in London, Paul Cambon, was a friend of Governor Pallain and in his letters to the French Foreign Ministry, he emphasized the political leverage over the British that could be gained through the provision of French financial assistance to the BoE. ¹⁴ So, too, did France’s ambassador in Berlin. ¹⁵ Cambon hoped that British financial dependence would help to cement a full Anglo-French alliance. When the financial crisis peaked a year later, Cambon wrote that:

Instead of dealing directly with the United States, the Bank of France must agree with the Bank of England to maintain the monetary situation in both countries. But it is also important that this assistance take a more official form than last year, and that our Bank strikes a direct relationship with the Bank of England. The two governments thus acting together on this operation, it will be useful to demonstrate … that France provides England with services that no one else, and least of all Germany, is in a position to offer her. ¹⁶

Nor could the BoF ignore strategic concerns in practice, even if it could do so in its formal minutes. Government control over the BoF was tightened in the law of 17 November 1897, which provided for: new permanent credit facilities to the state; a new tax on the amount of notes in circulation; a new rule preventing the governor and his deputies sitting in Parliament so as to make these three appointees less likely to oppose the finance minister; and a new governor, with Joseph Magnin replaced by career civil servant Georges Pallain. The governor handled international affairs personally, though the Rothschild regent and his English brother still provided valuable information and help in liaising with the BoE. Such autonomy on international matters as the governor enjoyed was conditional on the maintenance of his domestic political and economic connections. This was even true, to some extent, for private French banks. As Feis (1930: 47) noted: “The French government stood guard over the nation’s savings, so they might be exchanged for favor or privilege, and the banks resembled embassies.”

Thus, at least from the time of the 1906–1907 crisis, the overriding diplomatic priorities of the French government — to secure and deepen its strategic relationships with Russia and Britain — provided additional reasons for the BoF to provide much more extensive and persistent financial assistance to London than it had given in 1890. This priority is overlooked by OEP approaches. Eichengreen (1992: 31) similarly mischaracterizes a “reciprocity norm” between central bankers in this period. BoF assistance by 1906 was given not in the expectation that the BoE would at some point return the favor, but because the French hoped that Britain would reciprocate with diplomatic and military support against Germany. Key figures in the British government also recognized this strategic logic. Sir George Clarke, Secretary of the Committee of Imperial Defence (established in 1902 to coordinate military strategy), wrote in November 1906 to Asquith (then Chancellor of the Exchequer) that “the entente cordiale is almost a financial necessity.” ¹⁷

As for the Reichsbank, its role in assisting Britain during the crisis was less extensive than the BoF’s. It permitted gold to flow to Britain and the US in 1907, but it did not discount British bills. Perhaps because of this more passive role, neither the minutes of the Reichsbank’s Directorate nor the foreign ministry archives mention any decision to allow gold reserves to flow to London in 1907 (though this may be because many German government documents were destroyed during the Second World War). There is circumstantial evidence supporting the claim that protecting domestic economic interests was a motivation for the Reichsbank in autumn 1907. Broz (1997) emphasizes the demand for low interest rates by the land-owning Junkers, while others argue that Germany’s rising industrial elite and its growing linkages to the major banks were the dominant interest favoring monetary stability (Gall, 1995).

However, these accounts ignore the German state’s changing interest as a borrower and its links with increasing international rivalry. Von Tirpitz’s costly naval expansion program, financed by increased state borrowing, caused yields on German government bonds to rise in the two years before November 1907 as financial circles became increasingly concerned about its cost (Ferguson, 1998: 129; Kennedy, 1980: 303, 358). The growing strain in Germany’s money markets over 1907 threatened to increase further the government’s borrowing costs, providing another reason for Reichsbank assistance to London.

Nevertheless, it would be wrong to assume that borrower demands for low and stable interest rates in Germany produced relatively stable Reichsbank discount rates compared to England. The Reichsbank raised the discount rate from 5.5% in mid-1907 to 7.5% on 11 November, the highest peak of the gold standard period for Germany. Furthermore, as Figure 1 shows, the volatility of German discount rates was significantly higher than in England and far higher than in France over the period 1900–1907. This suggests that the tolerance of the German economy for monetary instability was comparatively high. Although it is possible that Reichsbank discount rates would have been even more volatile in the absence of (limited) assistance to the BoE during crises, the data suggest that the prevention of discount rate increases was not an overriding consideration. Figure 1 also shows that after foreign financial assistance by Germany to England ceased completely after 1907, the volatility of Reichsbank discount rates actually fell slightly compared with the earlier period (discussed later).

OPEN IN VIEWER

Figure 1.

Variance in major central bank discount rates, 1900–1907 and 1908–June 1914.

As for France, international political and strategic considerations had become more important in Berlin by 1906. The Reichsbank was subject to even greater political control than the BoF. ¹⁸ Major German private banks were represented on its central committee but, as in France, they were close to key political figures and well-understood the complex linkages between finance and politics, both domestic and international (Gall, 1995; Stern, 1979). Any involvement of the Reichsbank in ILLR operations necessarily required governmental approval.

The Triple Entente between Britain, France, and Russia was far from a full-fledged military alliance, let alone a guarantee of British solidarity in the event of war with Germany. Even the more aggressive proponents of Weltpolitik surrounding the German Kaiser hoped that Britain’s imperial rivalries with France and Russia, never entirely resolved by the agreements of 1904 and 1907, might lead Britain to choose neutrality in a continental conflict. Financial circles in Germany were among the more pro-British elements of the German elite and lobbied the government actively after the 1905 Moroccan crisis to promote better relations with London (Kennedy, 1980: 47–48, 302–305). Von Bülow, then Germany’s chancellor, was also very conscious of the political consequences of permitting France to wield its unrivalled financial power in Europe (Ferguson, 1998: 39). Allowing gold to flow to London during the 1907 crisis was therefore consistent with Berlin’s desire to discourage the emergence of a more solid British–French–Russian pact.

In summary, the protection of domestic economic interests from interest rate volatility, as emphasized by OEP scholars, provided one motivation for French ILLR assistance to Britain over 1906–1907. However, there are also strong indications that strategic considerations were playing an increasingly important part in French financial decision-making by this time. The overriding diplomatic objective of building support for French interests in Europe provided an additional and increasingly powerful reason for using financial policy to deepen British dependence on Paris. France’s central role as an ILLR was becoming overdetermined: both domestic and international strategic factors now shaped its central role as crisis manager. As for German policy over 1906–1907, the available evidence is very limited. It is plausible that German policymakers believed that financial assistance to Britain would benefit domestic societal interests, but German policymakers also hoped that such assistance would benefit the state, both financially and strategically. Thus, the OEP approach provides a parsimonious explanation of the behavior of central banks during the 1906–1907 crisis, but it ignores the growing impact of an increasingly politicized and securitized environment on key decisions in Paris and Berlin.

The role of security factors in shaping international financial policies by this time was even starker in a case that has received too little attention in the IPE literature. On the brink of financial collapse, Russia successfully floated in April 1906 a massive bond issue in Paris that raised 2,250 million French francs. This was the equivalent of £90 million — a sum that dwarfed the international financial assistance provided to the BoE in the 1890 and 1906–1907 crises. Although the BoF itself did not provide the funds, the loan was strongly supported by the French political and financial establishment and served state interests — after all, the future of the Dual Alliance was at stake. The British government also recognized its strategic value and London bankers were closely involved. Almost certainly, the loan was a substitute for more direct official financial support to St Petersburg. Sergei Witte, the former Russian finance minister, architect of Russia’s adoption of the gold standard in 1897, and the 1906 loan’s main negotiator, regarded it as “the loan that saved Russia” (Wcislo, 2011: 232; Yarmolinsky, 1921: 285–315).

The Russian financial and political crisis of 1905–1906 was a direct consequence of Russia’s disastrous war with Japan in 1905. Russian state finances had collapsed, gold reserves fell sharply and the government needed a large international loan to avoid default and the abandonment of its gold peg (Crisp, 1961: 499–500; Neilson, 1995: 102; Yarmolinsky, 1921; Ziegler, 1988: 312–315). The Russian Treasury consulted in late summer 1905 with French and British bankers about a large loan. One of the Barings brothers, Lord Revelstoke, immediately asked Lord Lansdowne, then British foreign secretary, if he would support it. Lansdowne indicated his support for such a loan due to his desire for a strategic rapprochement with Russia (Neilson, 1995: 102). Negotiations suffered a series of delays due to growing domestic political unrest in Russia, continuing Russian suspicion of Britain because of its alliance with Japan, and the election of a Liberal government in Britain in January 1906, which raised doubts about whether London could support the financing of an autocratic government.

However, Sir Edward Grey, the new foreign secretary, told Lord Revelstoke privately that “while from the financial point of view,” Grey could not “take the responsibility of tendering any advice … from a political stand-point he would view with satisfaction” the conclusion of the loan (cited in Neilson, 1995: 273). Like his Conservative predecessor, Grey avidly favored a broader strategic rapprochement with Russia. The British Foreign Office was also concerned that a Russian financial collapse would create turmoil in France and weaken Paris’s position in the Algeciras conference. For all of these reasons, the Foreign Office congratulated Barings on its contribution to good Anglo-Russian relations on the conclusion of the loan (Neilson, 1995: 103, 271–272).

Domestic commercial interests in France were not pushing strongly for a new loan to Russia, as the OEP framework might predict. In fact, the French government vigorously encouraged very reluctant French bankers to extend more credit to St Petersburg to support their ailing ally. As in Britain, the radical government elected in March 1906 continued this policy despite the domestic political difficulty of assisting a foreign autocrat. In the end, finance was one of France’s key strategic levers and the temptation to use it was great. The French government used delaying techniques to force Russia to conclude peace with Japan, to bind Russia more closely to France and to resist German overtures, and to encourage an Anglo-Russian rapprochement (Crisp, 1961: 500–508; Neilson, 1995: 273). In all of these respects it was largely successful.

Even Germany, increasingly isolated by this Anglo-French diplomacy, initially thought that it might be able to exploit Russia’s desperate need for finance to obtain its support in the 1905 Moroccan crisis (Joll and Martel, 2007: 196–197). However, Germany lacked the financial resources of London and Paris and its broader bid to wean Russia from the French orbit was spurned. In reaction, the Berlin government vetoed further German bank loans to Russia (Feis, 1930: 172; Viner, 1928a: 166–168). Witte himself described this veto on German bank participation in the Russian loan as “treacherous” and as “an act of vengeance for Algeciras and for our rapprochement with England” (Yarmolinsky, 1921: 304).

Other historical studies show that the 1906 loan to Russia was exceptional only in size. Private international finance increasingly “followed the flag” in the decade before 1914. ¹⁹ In the largest case of international lending to a crisis-hit European government before 1914, security concerns had become paramount for official actors and the OEP framework has limited utility.

Empirical and theoretical implications

In opting for a methodological reductionism that sets aside international factors as determinants of foreign economic policy, the OEP approach risks overlooking the potential dependence of such policy choices on the macro-level environment (Oatley, 2011). More specifically, I have argued that as international systems become more rivalrous and insecure, national security considerations will play a larger role in shaping international financial policy decisions. In such circumstances, the appropriateness and utility of OEP approaches will diminish.

I have shown that this was true under the pre-1914 international gold standard, when rising levels of insecurity increasingly shaped the level and nature of international lending during financial crises. This is at odds with the consensus view in the IPE literature, which sees such lending as a consequence of a fortuitous coincidence of domestic interests and institutions. The consensus is broadly accurate in the period up to about 1900, but not thereafter. European international financial policies were increasingly shaped in the new century by the objective of building and cementing alliances, and later by the desire to hoard gold in preparation for war. As we have seen, when a series of financial crises occurred from 1906, governments, central banks, and even private financiers were increasingly constrained by strategic considerations that played much less of a role in the 1890 Barings crisis. From 1906, British dependence on French financial assistance during periods of financial instability increased, fulfilling one key objective of French strategy, and Britain and France cooperated in the largest financial rescue of the whole period, the Russian loan of 1906, as part of a general drift toward an anti-German security policy. Germany’s international financial policy became increasingly volatile after the first Moroccan crisis, with its diplomatic disappointments leading it to opt out entirely of financial cooperation with its strategic rivals. All the major powers excepting Britain hoarded gold as the security situation continued to deteriorate after 1907. Strategic rivalry increasingly shaped and distorted the operation of the international gold standard in ways that are captured neither in economics texts nor in OEP accounts.

The strong implication is that the usefulness of the OEP approach is context-dependent: it produces accurate predictions in relatively tranquil international systems, and, at other times, only because domestic and international considerations are aligned. That this has been overlooked in the IPE literature on this period is probably, in part, because in the important case of France, the alignment of domestic and strategic motivations was so marked. Yet, even for France, security factors increasingly shaped foreign financial policy in ways that were not favored by domestic societal interests. The French government pushed reluctant bankers to lend enormous sums to a near-bankrupt Russian autocracy in its hour of greatest need, and the priority of building gold reserves reshaped the nature of the financial support that the BoF provided to London.

This interpretation is consistent with the more general point that international financial relations are inescapably embedded in and can be reshaped by an evolving international political order (Cohen, 1998; Helleiner and Kirshner, 2009; Kirshner, 1995, 2003, 2007). In focusing only on domestic structures, the OEP approach can too easily overlook this. My argument also suggests that the classical gold standard should not be seen, as it often is, as a single, homogeneous system over the whole period 1873–1914. De Cecco (1984: 120–126, 128ff) also argues that the classical gold standard had already begun to unravel in the decade before 1914 because the rise of the US as a major economy with a volatile demand for gold undermined Britain’s ability to retain its anchor role as the country with free gold convertibility. This may be true, but it also overlooks the importance of new strategic priorities on the European continent.

I do not contest the argument that systemic change can be produced by domestic-level change. Various scholars argue that after the First World War, the domestic political foundations of the gold standard were eroded by the extension of the political franchise in major countries and as working-class organization increased (e.g. Gilpin, 1987: 126; Simmons, 1994: 20–29). This should not prevent us, however, from recognizing that the international system can also produce policy change. My analysis shows that it is misleading to distinguish between a “depoliticized” international gold standard during the 1873–1914 period and a highly politicized interwar gold standard (Kirshner, 2003: 269; Simmons, 1994: 30; Toniolo, 2005: 13–19). Since international political rivalries continued to shape central bank relations after 1918, there is more continuity between the two eras of the gold standard than is commonly recognized (Ahamed, 2009).

Turning to more recent applications, I have argued that much higher levels of international institutionalization since 1945 are also likely to increase the importance of international factors in foreign financial policy decisions and to diminish the usefulness of the OEP framework. In highly integrated and deeply institutionalized systems such as the EU, states are more prone to devise policy responses within shared institutions in which policy norms are generated and diffused, even if these norms are contested. This also seems to be true at the global level, as the coordinated response of major governments to the 2007–2009 crisis through the Group of Twenty (G20), the IMF, and other international institutions suggests.

Yet, it is still possible to see the influence of changing levels of international rivalry and security on financial policy decisions in recent years. International emergency lending during the Cold War era — both through the IMF and within narrower groupings such as the Group of Ten — was deeply embedded in the political and military relations of the Western bloc (Thacker, 1999). Since 1990, emergency international lending has operated in the shadow of a rapidly evolving international political system. Contrast, for example, the substantial IMF assistance to the Yeltsin government of Russia in the 1990s with the Western financial sanctions imposed on the Putin government following the Ukraine crisis of 2014. Consider, too, the rival efforts of Russia and the West to lend to Ukraine in order to prise it from the other’s orbit. ²⁵ Nor has the pro-Moscow tilt of the new Greek Syriza-led government increased the willingness of other EU governments to compromise in ongoing debt negotiations. ²⁶

Ignoring the role of the international security environment in shaping such policies is especially problematic in an era when new powers are challenging the community of advanced democracies. My analysis suggests that we should be cautious regarding the suitability of the OEP framework to explore effectively issues as diverse as the redistribution of votes and quotas in the IMF, regional currency cooperation, bilateral currency swaps, and development finance initiatives. ²⁷ Demands by emerging countries for more influence within existing international financial institutions, and implicit threats by China and other BRICS (Brazil, Russia, India, China, and South Africa) countries that they can build alternatives to them, also show that international political dynamics are reshaping the provision of official international finance. It also suggests that the international security literature should explore more systematically the relationship between national security and international financial and monetary policy.

Finally, my argument supports the claim that the turn away from systemic theories in IPE in recent decades has had significant costs (Cohen, 2007; Keohane, 2009; Mastanduno, 1998: 853; Oatley, 2011). Of course, it would be wrong to opt for a wholly systemic approach: I have argued that domestic factors are often important determinants of foreign financial policy, but their degree of importance is contingent on international system dynamics. The shadow of insecurity and war looms much larger in some periods and contexts than in others, even during periods associated with rapid economic globalization, as after 1890. Kirshner (2003: 3) is surely right to argue that “monetary phenomena are always and everywhere political,” but the nature and degree of this politicization varies substantially over time.