Introduction
The 2007–9 financial crisis and subsequent bailouts drew new attention to the power of banks. In particular, the “too-big-to-fail” narrative highlighted how states’ need to preserve a functioning financial system meant they needed to protect their banks. The crisis also spurred a wave of domestic and international reforms of financial regulations, of which the centerpiece international effort was the reform of the Basel Accords that came to be known as Basel III. This article connects these two threads by examining how the relative structural power of banks in different financial systems shapes the speed of implementation of Basel III through a quantitative analysis.
This article contributes to both the structural power literature and the literature on Basel III. Firms possess structural power if they play such an essential role in the operation of the economy that policy makers act to protect the firm's interests even in the absence of lobbying, political pressure, or other forms of instrumental power. Examining structural power presents substantial challenges. If policy makers act in the interest of firms, it may not be clear if this is because of the structural importance of the firm or the instrumental power of the firm's direct efforts to shape policy. However, cross-national comparison of common pressures offers a way to compare structural power across systems. The 2007–9 financial crisis revealed that the enduring greater structural power of banks in bank-based systems versus capital market-based systems shapes state policy in response to the common shock of financial crises. This article expands on earlier insights in the structural power literature by exploring the degree to which that structural power also manifests in response to the common shock of international regulatory changes such as Basel III.
Most work on Basel III has either focused on the negotiation of the Accords and modeling the expected impacts of its components or on case study analysis of the implementation debates in a handful of countries. There has only been limited quantitative analysis of implementation, reliant on states’ voluntary reporting. This article supplements that earlier work by using implementation reports from the Bank of International Settlements (BIS) to parse true compliance from “mock” or “idiosyncratic” compliance that does not genuinely conform to states’ international commitments. Moreover, it links the Basel III debates with the structural power debate by identifying clear patterns in implementation of the Accords based on the structure of national financial systems, as well as national wealth and political structure.
This article will proceed as follows. The first section reviews the literature on structural power, especially in financial systems, and discusses strategies to separate measurement of structural power from instrumental power. The second section reviews the literature on the Basel III Accord and discusses the expected impact on banks of the major elements of the Accord. The third section lays out a series of hypotheses on how the structural power of different national financial systems is expected to shape domestic implementation of the elements of Basel III. That is followed by a discussion of the dataset and model used to test these hypotheses. The article concludes with a discussion of the results of those regressions.
Literature review
How to account for the structural power of firms in capitalist states is an old question, but one that acquired renewed interest in the aftermath of the 2007–9 crisis. The modern conversation largely starts with Charles Lindblom's argument that states in capitalist economics rely on firms and markets to allocate material resources and therefore need firms to thrive for the economy to function.Footnote 1 This therefore gives those firms structural power; independent of any lobbying by firms, states will act to ensure their firms’ health to preserve a well-functioning capitalist economy. Culpepper updates this concept by defining structural power as a set of mutual dependencies between business and the state, providing a useful baseline definition for contemporary discussions of the topic.Footnote 2
While Lindblom's work primarily focused on divisions between capitalist and communist economies, much fruitful work since then examines differences in the structure of financial systems within a capitalist context. In particular, the capital market-based/bank-based dichotomy first outlined by Zysman highlights the relatively greater structural power of banks in states where banks dominate provision of capital.Footnote 3 A healthy financial system, and thus healthy financial firms, are essential to the state in both types of systems. However, state economic policy must be sensitive to the specific nature of the national financial system.Footnote 4 In states with large capital markets, firms can raise money through stock and bond markets, making healthy banks less essential than in bank-based systems, where banks are the predominant source of capital and credit. Therefore, states must be more sensitive to the needs of their banks in bank-dominated systems.Footnote 5
Global financial integration and increased reliance on market financing in both kinds of systems in the early-twenty-first century led to the rise of the convergence thesis, arguing that the diversity of systems identified by the comparative capitalisms literature would soon be flattened, diminishing its importance in economic policy making.Footnote 6 However, the impact of convergence has been to date at a minimum overstated. Banks continue to play a central role in the allocation of capital in bank-based systems even with an increase in market-based financing, and state policies continue to diverge to accommodate the structural differences in their financial systems.Footnote 7 The 2007–9 financial crisis in particular revealed that structural differences in national financial systems continue to shape state policies in substantial ways. Mitchell and Woll both found substantial differences in policy responses between states with capital market-based financial systems and those with bank-based systems.Footnote 8 Woll found that bank-based systems produced more private collective responses, while Mitchell found that when state funds did become involved, states with bank-based systems issued more generous public bailouts. Culpepper and Reinke, meanwhile, argued that bank reliance on a single domestic market gives that state greater leverage over banks to enforce more stringent bailout terms.Footnote 9
Measuring structural power and its impact presents substantial challenges. A measure of structural power must be independent of a measurement of its effects or risk a tautology.Footnote 10 Scholars also divide as to whether to focus on the structural power of individual firms or a sector as a whole. The individual-firm approach emphasizes the diversity of interests that may exist among banks within a common system, especially how the interests of the most prominent or internationally active firms within a system may dominate financial politics.Footnote 11 The sectoral approach emphasizes the functional needs of the economy as a whole, and thus the state's need to maintain the health of the entire banking sector regardless of the fate of individual firms.Footnote 12 Scholars using the latter approach have focused on building a measure to separate the different models of national financial systems, which can then be used to examine the different systems in comparative perspective. Mitchell and Copelovitch and Singer both use a ratio of market capitalization to private bank assets to accomplish this, though this measure cannot capture intrabank divisions within a national financial system.Footnote 13
Structural power must also be untangled from instrumental power.Footnote 14 This presents two difficulties. The first is the mutually reinforcing nature of the two forms of power. Instrumental power can be used to bolster structural power by changing or protecting the institutional status quo, and structural power may cultivate the instrumental power of structurally important firms by increasing profits that can be used instrumentally.Footnote 15 This entanglement makes it hard to determine whether structural or instrumental power drives bank influence, especially as banks engage in substantial lobbying and other application of instrumental power regardless of their structural power.Footnote 16 Secondly, while the structural power of individual firms may change as they grow or shrink, the structural power of the sector as a whole varies only slightly over time.Footnote 17 For instance, the structural centrality of German banks, established in the late nineteenth century, has survived three regime changes, defeat in two world wars, and the Great Depression.Footnote 18
Two potential strategies offer promising routes to examine structural power and its impact. The first is careful process-tracing analysis sensitive to the difference between structural and instrumental power, such as Fairfield's examination of tax policy in multiple Latin American states.Footnote 19 Even here, however, separating the effects of instrumental and structural power may be challenging at best and subjective at worst.
The second approach is cross-national comparison. Structural power may vary relatively little within a single case, but can vary substantially across countries. Effective use of this approach requires a common pressure on the different systems, such that differences in outcome cannot be simply ascribed to the differing nature of the challenges presented. This has made study of exogenous shocks, such as global financial crises, tremendously valuable, as they present a common pressure on multiple systems that can reveal key differences ascribable to structural power differences. Such studies may also include comparative examination of the postcrisis reform agendas in crisis-struck states.Footnote 20 However, as a crisis recedes and country-specific factors grown in importance, it becomes harder to directly compare domestic reform agendas.
Adoption of new regulation through international agreement presents another common pressure that can be used as a basis for cross-national comparison of structural power. As in a financial crisis, the states and their financial systems all face a common shock, enabling a comparison which controls for the nature of the challenge facing the system. Such international regulatory moves could be dubbed “endogenous shocks,” as changing rules of international integration reflect policy choices by relevant states and other actors to set the “rules of the game” in a particular way.Footnote 21
Much of the structural power literature considering international agreements focuses on the negotiation of the agreement. The literature on structural power and negotiation of international agreements here splits between international political economy (IPE) approaches focused on the structural power of key states and comparative political economy (CPE) approaches examining the power of transnational firms. IPE approaches generally take “most powerful states” to mean the states with the largest financial markets, especially the United States and, to the degree it acts as a unified actor, the European Union, though some scholars emphasize questions of access and technical expertise, where again the great powers should dominate due to their superior resources.Footnote 22 CPE approaches focus on private actors using international fora to promote their interests. CPE approaches from the historical institutional tradition emphasize how domestic organization of interest groups and the structural power this provides them shape international negotiations, while exploring how international regulation can shape domestic interests and structural power in a policy feedback loop.Footnote 23 The New Interdependence Approach adds a focus on how actors able to draw on transnational networks of allies can use international regulation to gain advantage over their domestically focused rivals.Footnote 24 This approach therefore encourages examining international regulatory arrangements at the global level, rather than focusing on country-by-country domestic dynamics.
Although much of the literature has focused more on negotiation, examination of the domestic implementation process once the states agree to terms at the international level may reveal more about the structural power of the banking sector as a whole, as opposed to individual firms. Only some public and private actors can effectively influence international negotiations, and those actors may have substantially different preferences from other domestic actors.Footnote 25 The impetus for transnational regulation may in fact be a strategic choice to use the limited access to transnational venues to end-run around domestic opposition.Footnote 26 Moreover, as the IPE literature emphasizes, outcomes from international negotiations may reflect the relative power of states, rather than the domestic structural power of banks relative to other domestic forces. The structural power of the banking sector as a whole, therefore, may only manifest at implementation, when interstate power dynamics and the instrumental power advantage of international firms diminishes with the shift to domestic venues. The shift to domestic implementation lowers barriers to access for interested parties, which no longer need the capacity to organize transnationally to engage in the debates.Footnote 27 It also potentially increases issue salience, as debate moves out of relatively closed international venues to domestic ones, making it easier for both supporters and opponents of the reform agenda to rally allies to their cause.Footnote 28
As noted preceding, international agreements creates a common shock enabling cross-national comparison of structural power. While all states that are party to the agreement agree to adopt the same regulatory approaches within the discretion allowed by the agreement, the domestic structural power resources of banks may vary considerably. In the absence of large capital markets, banks play a central role in capital allocation, which in theory makes states more sensitive to their structural importance. While research on financial crises has demonstrated that the bank-based/capital market-based dichotomy carries explanatory power in the face of a common exogenous shock, the explanatory power of that dichotomy in the postcrisis period has been relatively underexplored. Examination of implementation of Basel III offers a chance to advance the comparative capitalisms and structural power literatures by examining the impact of a common pressure not directly linked to the financial crisis. Looking at implementation of the most significant common postcrisis international regulatory policy across the Basel member states should reveal the degree to which the relative structural power of banks shapes financial policy outside of a crisis moment.
Tensions in implementation may be particularly acute in “second mover” countries that failed to win their preferred outcomes in the negotiation phase, but nevertheless signed on to the ultimate agreement. As Drezner notes, smaller states frequently find themselves rule-takers to the great powers.Footnote 29 The financial great powers may also be rule-takers; if the European Union and United States disagree on a regulatory approach and compromise is not possible, one or the other must accept defeat. Thus, the United States found itself on the losing end of several policy debates in Basel II, and the European Union on the losing end in several Basel III debates, putting each in the position of agreeing to implement an international agreement against its own initial stated preferences.Footnote 30
Basel III
The Basel III Accords provide a key opportunity to explore the structural power of bank-based and capital market-based systems in response to common international regulatory pressures. Now in their third major revision, the Basel Accords are the preeminent global agreement setting standards for banking regulation. They establish a global minimum for financial regulatory standards negotiated by the Basel Committee on Banking Stability (BCBS). The Basel Accords provide an ideal case to study implementation dynamics in large part because they are soft law. The Accords establish common standards and the BIS monitors compliance, but neither the BIS nor the BCBS can formally censure members beyond “naming and shaming.”Footnote 31 Although some fear that such a soft law approach may impede the effectiveness of the Basel Accords,Footnote 32 this does not appear to have been the case overall.Footnote 33 The Basel Accords are a lynchpin of global financial regulation, and rejecting them would carry great costs both in terms of reputation and access to foreign markets. States failing to meet their Basel commitments may find their access to foreign financial markets curtailed.Footnote 34 Private actors also may shun a noncompliant state, creating market pressure to comply.Footnote 35 A soft law approach and lack of enforcement mechanism does, however, make implementation of Basel III ripe for strategies of delay and noncompliance.
While outright rejection of Basel III commitments may be too costly to consider, several other options to avoid or delay compliance may remain. One option is what Walter dubs “mock compliance,” enacting statutes that on paper conform to international agreements, but that go unenforced in practice.Footnote 36 A more sophisticated strategy may be dubbed “idiosyncratic compliance.” A state may adopt and enforce a rule stated to comply with international commitments that in its actual details diverges substantially from the substance of those commitments. Critics alleged that the European Union's Capital Requirements Directive IV (CRD-IV) did just that. It theoretically brought the EU into alignment with Basel III standards but some saw it as so divergent as to border on noncompliant.Footnote 37 Finally, a state may avoid implementation of undesirable global standards through delay, never formally abandoning a commitment to embracing the standards but pushing off their implementation. This strategy in some ways may be the most effective. While international monitoring may expose mock or idiosyncratic compliance, delay can be simply blamed on a slow-moving policy process. At its most successful, this strategy can prevent implementation of undesirable standards until they are renegotiated internationally while avoiding ever formally breaching the accord. The United States successfully adopted this approach with elements of Basel II, delaying implementation until changes under Basel III brought international standards more in line with American preferences.Footnote 38
Literature to date on Basel III implementation has generally pursued one of two tracks. The first focuses on case study examinations of single country implementation of the accords. Unsurprisingly, the most commonly examined cases are the United States and the European Union, the world's two largest financial systems.Footnote 39 Scholars have also examined implementation in Australia, India, Turkey, and others.Footnote 40 These studies generally focus on identifying the specific costs to the domestic banking sectors and their capacity and strategies to resist or reduce them. Apart from some limited comparison between the United States and the European Union or among European Union member states, they generally consider these cases in isolation from each other, making it difficult to separate the structural and instrumental power of actors.
The second major track examines the economic impact of Basel III, either in toto or focused on some specific component. Much of this work models anticipated impacts, work that feeds into the debates on implementation by shaping perceptions of anticipated costs and benefits.Footnote 41 As implementation progresses, an increasingly prominent thread in the literature examines the early actual results of implementation, either in specific cases or globally.Footnote 42 In general, this literature finds a likely increase in stability with a moderate increase in costs and complexity. However, it generally treats implementation as a given, and does not explore the politics around the “download” of Basel III commitments to national regulation.
Systematic examination of the political dynamics of Basel III implementation has been limited. While some of the economic impact-focused work offers quantitative measures of the economic effects of Basel policies across countries, Jones and Zeitz offer the only prominent quantitative analysis of the politics of implementation.Footnote 43 This article complements their work by examining a smaller group of countries in greater detail, as it relies on the independently verified survey in the BIS Regulatory Consistency Assessment Programme rather than the broader but voluntary Financial Stability Institute (FSI) surveys. Although this limits the universe of cases to BCBS member states, it also makes it possible to distinguish true compliance from mock or idiosyncratic compliance, which is not possible in the voluntary FSI surveys.
Banks rarely supported or opposed Basel III in its entirety, but rather supported or opposed individual rules based on their particular impact.Footnote 44 The new rules in Basel III can be broadly divided into three categories for the purposes of this analysis based on whether they were anticipated to raise bank costs, lower bank costs, or not have a substantial impact. What matters here is the impact anticipated by banks and regulators, not the actual effect. Actual costs may end up being higher or lower than anticipated but using actual costs to explain the politics of implementation would be post hoc reasoning. For instance, many observers expected the new liquidity rules to increase costs to banks, but multiple surveys found an insignificant impact in practice.Footnote 45
Perhaps the most prominent rule expected to increase costs was the countercyclical buffer, which increased Core Tier 1 capital ratios from 2.5 to 4.5 percent.Footnote 46 This rule was at the center of Basel III's move to increase bank resilience to crisis, though critics argued that its approach would be too insensitive to country-to-country differences and would raise costs and disincentivize lending without increasing stability.Footnote 47 A host of other risk-based capital measures were also introduced, though as discussed in the text that follows, many of these rules were anticipated to potentially lower bank costs, offsetting the costs of the countercyclical buffer and achieving the Basel Committee's goal of making Basel III essentially cost-neutral for fully compliant banks.
The second major reform was the introduction of leverage ratio requirements. The core capital requirement is sensitive to the risk-weighting of a bank's assets, but the 2007–9 crisis demonstrated that even banks with a low-risk pool of assets could be vulnerable if they were highly leveraged. The new leverage ratio added a second capital ratio that simply weighs the value of a bank's capital to its assets regardless of asset quality. It thus potentially raised costs for banks that had previously relied on a low net asset risk-weighting to minimize their capital requirements but now would still be required to increase their capital to meet leverage requirements.
The third major reform introduced new global common liquidity standards, out of recognition that the liquidity freeze in the 2007–9 crisis drew otherwise healthy banks into distress. Therefore, the Liquidity Coverage Ratio (LCR) created minimum standards for banks to hold cash and other high-quality liquid assets to deal with short-term liquidity crunches and the Net Stable Funding Ratio (NSFR) addressed longer-term liquidity access by requiring a large pool of available stable funding sources. Because these were new and substantial requirements for banks, they would prove to be among the most contested reforms by banks seeking to avoid new regulatory burdens.Footnote 48
Finally, Basel III attempted to address the “too-big-to-fail” problem by introducing new rules for banks determined to be systemically important at either the domestic or global level. These new rules were anticipated to substantially increase costs for domestically- or internationally-systemically important banks (D-SIBs and G-SIBs, respectively) by raising their capital requirements another 1–2 percent.Footnote 49 They also would reduce the implicit subsidy priced in by lenders for “too-big-to-fail” status by about half.Footnote 50
A series of other rules were anticipated to lower the costs of banks when fully implemented. The new capital requirements for central counterparties (KCCP) and securitization framework nominally increased burdens on banks, by increasing the risk weighting of assets cleared through central clearinghouses and increasing risk sensitivity for weighing of securitized assets respectively. However, full adoption of both offered opportunities for banks to lower, rather than increase, their required capitalization. Under the new KCCP standard, assets cleared through a Qualified Central Counterparty (QCCP) have a risk weight of only 2 percent. Therefore, implementation of the new rule and designation of such QCCPs allowed banks to reduce the cost of their assets in their risk weighting.Footnote 51 Similarly, full implementation of the securitization framework includes lower capitalization requirements for compliant banks.Footnote 52 Therefore, banks had reason to favor both measures, as they could help offset the capital increases elsewhere in Basel III if fully implemented.
The new standardized approach for measuring counterparty credit risk exposures (SA-CCR) was something banks actively sought out of the Basel III negotiations. It replaced an earlier measure, the current exposure method (CEM), which was seen as too crude, as it treated all credit swaps essentially identically regardless of time duration.Footnote 53 Similarly, the new rules on capital requirements for equity investments in funds replaced the old Standardized Approach's blanket treatment of equity investment in funds at 150 percent of risk-weighted assets with the option of more granular look-through approaches, enabling banks to potentially lower the risk-weighting of their investment portfolios by being more sensitive to individual fund risks.Footnote 54 Banks therefore would want to move toward these new and more sensitive measures as soon as possible.
Despite the initial predictions of some observers, banks also enthusiastically embraced the new total loss-absorbing capital (TLAC) rules. The TLAC rules, which only apply to systemically important institutions, set up a framework to allow for “bail-in,” where subordinated bonds convert to equity if a bank is in crisis.Footnote 55 In theory, this pushes the burden of saving a failing bank away from states and onto creditors. In practice, it allowed banks to raise capital through contingent convertible bonds and other convertible instruments at a lower price than conventional equity, arguably because lenders fail to appreciate that they're functionally providing equity rather than credit.Footnote 56 Enactment of the new TLAC rules therefore offered systemically important banks a new and lower-cost way to raise equity, and they eagerly sold billions in subordinated bonds as soon as the option became available.Footnote 57
Finally, the new interest rate risk in the banking book (IRRBB) standards provided new guidelines for how regulators manage risks to banks from interest rate changes. These changes include both maturity challenges of long-term maturity assets funded by short-liability assets and rate mismatch challenges of fixed-rate loans funded by variable-rate deposits.Footnote 58 The new rules established a required framework for banks to manage these risks. They were not anticipated to be clearly superior for banks from a cost standpoint, but their implementation was described as a “laboratory for testing on a small scale how prepared an institution is for putting into practice the big ideas that underpin” the rest of Basel III's new risk management and reporting regime.Footnote 59 To that end, banks had an incentive to implement them sooner rather than later, as a “test run” for later elements of Basel III.
Basel III also included a series of new rules on disclosure. These would increase the transparency of bank operations and complement the changes discussed but were not anticipated to have a substantial direct cost to banks. As such, they are not examined here.
Hypotheses
Given the expected impact of elements of Basel III and theory laid out the preceding text, the following hypotheses can be formed:
• H1: States with financial systems dominated by bank provision of capital will be slower to implement those Basel III measured expected to raise bank costs and faster to implement those measures expected to lower bank costs.
Banks in both types of systems obviously have an incentive to oppose measures that are seen as likely to increase their costs, and to support measures to lower them. However, banks should have less structural power in states with large capital markets. Because other market actors can turn to capital markets for funding, banks are less crucial to the functioning of the overall economy. Therefore, policy makers can more easily disregard the preferences of the banks in favor of stability enhancing measures that will benefit the overall economy. By contrast, in states where banks dominate financial systems, states will need to be more responsive to the immediate needs of banks, and thus more likely to accede to delaying measures that will increase bank costs and accelerate implementation of measures that will lower them.
• H2: States with greater international financial integration will be faster to implement all Basel III measures.
One of the key purposes of the Basel Accords in all their iterations has been a harmonization of international financial standards. While this at its core is about avoiding a “race to the bottom” in financial regulation, harmonization of regulatory standards may also serve the interests of transnational banks. Harmonization lowers transaction costs for cross-border operations and helps eliminate national rules protecting domestic banks against foreign competition.Footnote 60 Moreover, implementation of the Basel Accords and other international financial regulatory standards can signal to foreign investors that a country has a mature financial system worthy of attracting foreign investment. Therefore, the more globally integrated a state's national financial system is, the more it should want to remain at the cutting edge of global financial integration, and thus push for rapid implementation.Footnote 61
• H3: Wealthier states will be faster to implement Basel III measures.
Although the cumulative impact of Basel III was intended to have a negligible net effect on costs to banks, it was widely anticipated to increase them.Footnote 62 Wealthier states, being in a better position to accept short-term costs to their financial systems in exchange for longer-term benefits from increased stability, should therefore be in a better position to implement Basel III.
The structure of national political systems should also shape implementation of international agreements. As Putnam traced in his work on two-level games, international ratification must be followed by domestic approval.Footnote 63 The more veto points that exist in a national political system, the slower implementation will be, and the more opportunities opponents of the agreement will have to delay or modify it to avoid costs to themselves.Footnote 64
• H4: Democracies will be slower to implement Basel III measures than authoritarian states.
Making substantial changes in a democracy requires assembling a broad coalition in support, meaning that autocracies should be able to move faster and with less opposition.Footnote 65 This is especially likely to be the case in regard to international financial agreements, where autocracies may face concerns over rule of law that could deter foreign investment. Regulatory harmonization through rapid implementation of international standards can therefore be valuable as a signal of the credibility and maturity of a state's financial regulator, and a useful tool to attract foreign investment, which should also speed adoption of Basel III rules.Footnote 66
• H5: States with presidential systems will be slower to implement Basel III measures than states with parliamentary systems.
Dividing policy making between independent executive and legislative branches obviously creates additional opportunities for friction and veto points.Footnote 67 Thus, greater delay or compromise in implementation should be expected in the “downloading” of international standards. This is especially likely to be the case in presidential systems where negotiation of international agreements is the sole preserve of the executive branch and the legislature has a direct role only in implementation of agreements, but not the drafting of them.
• H6: States with proportional representation electoral systems will be slower to implement Basel III measures than states with single-member district electoral systems.
Proportional representation ensures that a wider array of viewpoints are represented in the legislature and typically also makes development of a multiparty system more likely. This, in turn, should give banks or other financial interests more opportunities to block implementation of unfavorable Basel III provisions.Footnote 68
Dataset and model
This article considers implementation of the Basel III Accords in all BCBS member states, excluding the EU as a separate entity, from 2011 to 2019 using data from Bank of International Settlements’ biannual progress reports on adoption of the Basel regulatory framework from the Regulatory Consistency Assessment Programme (RCAP).Footnote 69 Using this data necessarily limits the universe of cases to the BCBS member states, though it gives two advantages. The first is that, while other states frequently adopt the Basel standards, the BCBS members have made an explicit commitment to do so. Therefore, failure to implement explicitly breaks of a previous commitment, increasing the reputational costs of noncompliance more than in nonmembers that decline voluntary implementation.
The second advantage is that the BIS makes an independent judgment as to whether member states have fully complied with the Basel standards. Thus, this data separates true compliance from mock or idiosyncratic compliance. FSI reports, the principle alternative, capture a wider swath of countries, but rely on voluntarily reporting by states.Footnote 70 FSI reports therefore cannot distinguish true compliance from mock or idiosyncratic compliance.
In line with previous work on international regulatory compliance, this article uses a Cox Proportional Hazard model to estimate the effects of independent variables on implementation of the components of Basel III.Footnote 71 Cox models were developed out of medical trials, to determine the effect of treatments on patient survival rates where failure, typically death, removes patients from further observation. A Cox model is appropriate because it captures length of “survival” and not just whether a case survives or not. It also accounts for how cases drop out of the study once they “fail,” which in this context means implementing the observed Basel III reform. The reported hazard ratio indicates the percentage change in likelihood of “failure” (reform implementation) for each standard deviation of a variable at a given point in time. Therefore, a ratio greater than one indicates rapid adoption is more likely, and a ratio less than one indicates a greater chance of delay in adoption. Conventional standard errors are reported.
Different components of Basel III had different timelines for implementation,Footnote 72 though its soft law status meant some states would complete implementation either ahead of or behind that schedule at their own discretion. This analysis considers speed of adoption of individual components relative to other states implementing the same component, not relative to not relative to the target deadline or to other components of the Accord.
Financial structure is measured by a ratio of capital market value and private bond capitalization to bank credit.Footnote 73 This captures whether provision of financial services is dominated by banks or capital markets. As discussed in the preceding text, where banks dominate, states must prioritize the interests of their banks in implementing regulation more than in capital market-based systems. As a robustness check, regressions were rerun using a ratio of capital market capitalization to bank finance excluding private bond capitalization. Results were not substantially different and are not reported here.
Foreign exposure is captured in two measures, the natural log of the ratio of offshore bank deposits to domestic bank deposits and foreign banks as a share of total bank assets. The former captures the degree to which domestic banks are engaged in transnational banking, while the latter captures the degree to which foreign banks have a presence domestically. Both indicate the degree of international integration, with more integration expected to be associated with faster implementation of Basel III.
All data on financial system size except foreign bank presence are in constant 2010 US dollars and taken from the World Bank's Financial Structure Database, averaged over 1990–2010 to minimize distortion from the 1997–98 East Asian and 2007–9 transatlantic financial crises.Footnote 74 Foreign bank presence is taken from Barth et al., averaged from 1999 to 2011, the period under observation in that study.Footnote 75
Per H3, wealthier states should more readily adopt Basel III provisions. Wealth is captured by GDP per capita in constant 2010 US dollars and taken from the World Bank's World Development Indicators database averaged over 2001–11.Footnote 76
Finally, a series of political controls are included to capture differences in political structure. H4 expects authoritarian regimes to more readily adopt Basel III provisions than democracies. Regime is measured by the Polity II score of the relevant countries, averaged for 2001–11. H5 expects presidential systems to implement more slowly than parliamentary systems, as the division between the executive and legislative branches creates additional points of access. Similarly, proportional representation systems provide easier access for a wider range of voices than single member district systems, so H6 expects proportional representation elections to be associated with delay in implementation of the Basel III reforms. Political variables are taken from Beck et al. and averaged over 2001–11.Footnote 77
Results
The model was run across all BCBS member states across all elements of the Basel III reforms monitored by RCAP surveys except the disclosure rules, which were not expected to produce significant costs or savings to banks. Elements with statistically-significant relationships with the independent variables are reported in Table 1. Several elements produced results without significant relationships, and are not reported here. Most surprising among these are the lack of significant results for the new leverage and enough national divergence to reveal a statistically significant pattern.Footnote 78 Implementation of the leverage and NSFR rules, meanwhile, was targeted for January 2018, meaning even states eager to implement were not required to do so until late in the period under observation.Footnote 79 Regardless, these nonfindings warrant further exploration elsewhere.
Table 1. Regression Results.
^p < .15; *p < .10; **p < .05.
Note: Cox proportional hazard ratios with standard errors in parentheses.
Results on financial structure broadly confirm H1. Where banks dominate provision of capital, states more rapidly implement the cost-saving measures of Basel III and delay implementation of more expensive changes. States where banks dominate provision of capital are slower to implement the new countercyclical buffer rules, the change that would impose significant costs on the broadest swath of banks by raising the minimum core capital rules regardless of systemic importance, liquidity provision, or leverage ratio. Conversely, in line with expectations, states with bank-dominated systems are faster to implement reforms anticipated to lower costs to banks: the KCCP, equity investment, SA-CCR, securitization, TLAC, and IRRBB rules. However, the relationship is only significant at the 0.15 level for equity investment and securitization.
Measures of international exposure have very surprising findings. While H2 predicted that international exposure would increase the speed of implementation of Basel III components, actual findings find both measures of international integration associated with a slower implementation of several measures, all of which were expected to lower costs or otherwise be desirable to banks. It may be possible that the expected effect would appear in a broader dataset, as the BCBS member states are all already at a high level of international integration, meaning that any variation in international exposure starts from a very high floor. Nevertheless, this counterintuitive finding requires further investigation, possibly through case-study process tracing.
Wealth is associated with more rapid adoption of many Basel III measures, in line with H3. This is unsurprising insofar as a wealthier country can better afford the short-term disruption of changing a regulatory framework. Notably, wealth is the only variable with a significant effect on the new G-SIB and D-SIB rules. The relationship is only significant at the 0.15 level for G-SIB rules, but only some member states have globally systemically important banks, so the universe of relevant cases is smaller. However, there is no significant effect on the countercyclical buffer, the most expensive of the provisions under examination. This may be because the credit-to-GDP ratio used to calculate the size of buffer creates a correlation that obscures the wealth effect.Footnote 80
Authoritarian states more rapidly implement a number of measures, as expected by H4. This fits with a veto points story; authoritarian states with fewer avenues for opponents to block implementation should be expected to move faster than democracies. The specific measures authoritarian states implement more rapidly also fit the signaling story. Autocracy is only significantly associated with measures anticipated to lower bank costs, and not higher-cost measures. Because rapid implementation of any Basel III elements can signal a state's suitability as a place to do business, picking the “low hanging fruit” offers a way for authoritarian states to signal suitability at a minimal cost to itself.
H5 predicted that parliamentary systems should be faster to implement Basel III elements because the legislative-executive fusion reduces veto points that can slow the process. This hypothesis holds up for the most expensive measure examined, the countercyclical buffer, but parliamentary government is associated with slower implementation of the various cost-reducing measures of Basel III. While this suggests that H5 is valid for high-cost measures, the other relationships are unexpected. One possible explanation is that most parliamentary systems among the BCBS states are European and most presidential systems are in the Americas, suggesting that this may be a proxy for a regional effect, though adding a control for EU membership did not change results. Here again, future process-tracing work could be valuable.
Finally, while H6 predicted slower implementation in states with proportional representation elections, the expected relationship was only significant for the countercyclical buffer and TLAC rules, and only at the 0.15 level. That is in line with expectations, though not a robust finding.
Conclusion
Quantitative analysis of the implementation of major elements of the Basel III Accords reveals several noteworthy patterns of implementation linked to the financial and political structure of BCBS member states. Political structures have a significant and largely expected impact on implementation. More centrally to questions of structural power, whether a state has a bank-based or capital market-based financial system has a clear relationship with implementation of Basel III elements. Put simply, states with bank-based systems are significantly faster to adopt those measures likely to reduce bank costs, but to delay or otherwise avoid implementation of the most costly measure, the new countercyclical buffer rule. This fits with a story about the relative structural power of banks in the different systems. Banks in both kinds of systems will have similar preferences, and will use their instrumental power to lobby for their preferred outcomes. However, the superior structural power of banks in bank-based systems gives them an added advantage; the state must be more attuned to the needs of its banks to preserve the broader economy independent of lobbying. Thus the superior structural power supplements instrumental power to make bank-friendly results more likely. Overall, this therefore indicates that the bank-based/capital market-based dichotomy continues to hold explanatory power despite its relative lack of sensitivity to the diversity of views within individual banking systems.
Several other findings are also worthy of further exploration. Counterintuitively, international exposure was associated with delayed Basel III implementation, an unexpected effect that would benefit from both process-tracing analysis and examination in a broader dataset. Similarly, it would make sense that bank-based systems would delay implementation of the new leverage and liquidity rules, but no significant relationship was found. Potentially a dataset considering more countries or more time would reveal the expected pattern, but the available data does not support a conclusive finding. These findings here also suggest the value of future process-tracing analysis outside of the range of this article. Overall, these results highlight that the division between bank-based and capital market-based financial systems continue to produce significant differences in how states respond to external pressures, whether in the form of international crises or common regulatory policies.
