Financial Development and Financial Fragility: Two Sides of the Same Coin?

The paper puts forward three interrelated reasons that account for the recent breakdown of the finance–growth nexus, comprising the use (or abuse) of the relationship by policymakers, capture by ruling elites and institutional constraints. Evidence consistent with these hypotheses is provided from a recent European banking crisis. It concludes that in order to restore the ability of finance to promote growth, the influence of ruling elites on the rule of law, financial regulation and supervision needs to be minimized.


Introduction
There is little doubt that well-functioning banking systems can help promote economic growth and can facilitate access to financial services for broad sections of the population. Sound financial intermediaries are trusted by small savers and can, therefore, mobilize savings; in so doing, they automatically widen access to financial services. In addition, they provide the financial infrastructure for the payments system upon which day to day economic activity relies. Resilient ATM networks, efficient mobile and online payment services, card services and clearing systems are all important ingredients in facilitating transactions that support economic exchange among households, firms and government in any economy.
Moreover, well-functioning banks provide the infrastructure that is essential for firms engaging in international trade and facilitate the flow of capital across borders. Last but not least, they are able to effectively address asymmetric information problems inherent in credit markets and, by doing so, provide financing solutions to high-quality investment projects that contribute to inclusive economic growth. The ability of finance to underpin economic growth and to widen access to financial services is well known and relatively undisputed.
Reality, however, paints a rather different picture. Although we have some examples where finance has worked well for some of the time, we also have numerous failures, culminating in the global financial crisis of 2007-2008 from which many economies have yet to fully recover. By and large, financial sector policies, including those intended to promote the development of the financial system in LDCs and regulation in developed countries, have often failed to deliver stable banking systems. Even where full-blown financial crises have been avoided, financial fragility has led to economic stagnation and protracted balance sheet recessions.
Specifically, finance has not made much of a difference to the poorest countries in the world, such as those in sub-Saharan Africa (SSA). While in principle financial systems could have helped to lift millions of people out of poverty, SSA banking systems remain highly underdeveloped and unconnected to the process of economic growth. 1 In Asia, where financial deepening has been more evident in the last 30 years, liberalization was frequently followed by significant bouts of instability, including the East Asian financial crisis of the late 1980s. Even in countries that experienced high growth rates in the past, such as Japan in the post-WWII period and China more recently, banking sector weaknesses have been associated with the slowdown of economic growth. In the case of Japan, the economy stagnated for several decades as a result of banking sector fragility. In China, although not enough is known about the recent slowdown, our conjecture is that it is not unrelated to the large volumes of non-performing loans (NPLs) in the banking system and the over-indebtedness of the corporate sector, including in important industries such as steel.
In Latin America, banking systems have remained relatively underdeveloped and captured by ruling elites (Calomiris and Haber 2015). In the transition economies of Central and Eastern Europe, weak bank balance sheets, characterized by rising NPLs, are now a significant impediment to growth. 2 Last but not least, the subprime crisis in the USA, which transformed itself into a global financial crisis resulting in record bank failures in the USA and much of Western Europe, including Iceland, has shown that even developed financial systems have failed to deliver on many counts. In the last two decades, developed financial systems have not only failed to promote sustainable growth but have had widespread deleterious socioeconomic consequences with political ramifications. To start with, the large bailout bills to save banks in industrialized countries have been the main cause behind the deterioration of public finances and the increase in sovereign debt. This has, in turn, caused a fiscal squeeze often leading to austerity policies that have weighed heavily on aggregate demand and slowed down the recovery from the crisis, with many unintended social and political consequences. Moreover, the bursting of house price bubbles in the USA and many countries in Europe has caused balance sheet recessions from which many economies have yet to recover fully. Finally, the global financial crisis has exposed weaknesses in the euro's architecture that have threatened and continue to threaten its survival, something that could have unimaginable political consequences.
What, in effect, has happened over the last two decades is that banking systems in many regions of the world have become larger and at the same time more fragile and prone to crises. 3 As a result, financial development has, in effect, become almost synonymous with financial fragility-it is not an exaggeration to argue that recent experience shows that it is the other side of the same coin.
The purpose of this paper is to put forward three plausible, albeit interrelated, hypotheses why and how banking sector development has gone wrong. While we do not 'test' these hypotheses empirically, we analyse an important case study of a recent banking crisis in the euro area that is consistent with these hypotheses. Beyond our case study, the three hypotheses could form the basis of a new research agenda that we believe would be very fruitful.
The first hypothesis is akin to a Lucas critique. We believe that policies to promote financial sector development, including financial deregulation, backfired because the empirical relationships that may have existed in the data from the early 1960s to the mid-1980s were not structural ones; hence, when policymakers tried to use (or abuse them), they broke down. The second hypothesis is a political economy one. Ruling elites in many countries captured the process of financial development to further their own interests. In some cases, especially in some of the poorest countries in the world, this has meant financial underdevelopment but in many other cases this resulted in financial development of the 'wrong kind', e.g. regulatory capture and excess levels of household credit, reflecting over-indebtedness. The third and final hypothesis relates to institutional weaknesses. Many countries have failed to reach a minimal threshold level of institutional development, particularly in relation to the rule of law and the enforcement of contracts that are necessary to underpin healthy credit market development. While such weaknesses are almost self-evident in the poorest countries in the world, such as those in sub-Saharan Africa, they are often present in much more developed economies.
These hypotheses are put forward in ''Three Hypotheses''. ''Case Study: The Political Economy of the Cyprus Banking Crisis'' provides our case study relating to the recent banking crisis in Cyprus, which demonstrates that the three hypotheses are indeed interrelated. Specifically, that experience shows that power elites can benefit from both rapid financial development and weaknesses in institutions such as prudential regulation and rule of law. Although such a set-up is prone to financial fragility and crisis, the ruling elite can maintain its grip on power, even if it has to adapt to new circumstances.

Three Hypotheses
Finance and Growth: An 'Abused' Relationship?
In the 1990s and early 2000s, a large body of empirical research, much of which was associated with the World Bank, produced findings that suggested that bigger financial systems are better for economic growth. Summarizing this body of research, Levine (2003) states: '…countries with better-developed financial systems tend to grow faster-specifically, those with (1) large, privately owned banks that funnel credit to private enterprises and (2) liquid stock exchanges… The size of the banking system and the liquidity of stock markets are each positively linked to growth. Simultaneity bias does not seem to be the cause of this result'.
Much of this empirical research utilized cross-country regressions in which the dependent variable was the growth rate, averaged out over 5 years or more, included variables typically found to be significant in cross-country growth regressions, including initial per capita income, intended to capture convergence, secondary education and the ratio of government consumption to GDP. Financial development was often measured by the ratio of liquid liabilities and/or private credit to GDP. Although initial levels of the regressors were used to address potential endogeneity, these regressions were challenged on econometric grounds 4 and, also on whether they applied to LICs. 5 Notwithstanding the criticisms, which, in effect, were already hinting that the relationship between finance and growth was, at best, a reduced-form one, the policy agenda of the 1990s and early 2000s pushed for policies that promoted rapid growth in bank liabilities and private credit, often by liberalizing or deregulating financial systems and privatizing government-owned banks.
Rapid credit growth and expansion in bank balance sheets was, however, often followed by asset price bubbles, financial instability and crises. As a result, the empirical relationship between finance and growth all but vanished once the 1990s were added to the sample. Rousseau and Wachtel (2011), for example, show that the relationship between financial depth and economic growth alluded to by Levine (2003) has become extinct; they provide evidence to suggest that this development reflects financial instability and in particular the greater frequency of systemic crises since 1990. In a similar vein, Demetriades and Rousseau (2016) show that financial depth is no longer a significant determinant of long-run growth. Instead, they find evidence to suggest that certain financial reforms have sizeable growth effects that can be positive or negative depending on how well banks are regulated and supervised. These kinds of findings are certainly consistent with the Lucas critique: the relationship between finance and growth broke down because policymakers tried to exploit it, without paying sufficient attention to prudential regulation and supervision. It is, therefore, not surprising to observe that the mechanism which led to the breakdown was financial instability.
This would suggest that the finance-growth relationship that was detected in earlier data was a reduced-form one, which would, therefore, not allow reliable statistical inference and cannot be the basis for policy analysis. 6 The latter can only be based on structural relationships between the underlying variables, which should include the institutional underpinnings of financial systems, i.e. prudential regulation and rule of law.
By the time the 2000s were added to the data, the relationship was beginning to exhibit reversal, possibly reflecting use or abuse by policymakers: more finance was now shown to be associated with less growth. Some researchers (e.g. Arcand et al. 2015) argue that the relationship between finance and growth has an inverse U-shape and that there can, therefore, be 'too much finance', once a certain threshold of finance has been reached. However, an alternative and, in our view, more plausible explanation is that the relationship between finance and growth found in the data was never a stable one, because it was never a structural relationship.
Indeed, it is very hard to pinpoint any economic theory that suggests that the empirical relationship between finance and growth is anything other than a reduced form. In the neoclassical growth model that motivates most cross-country growth regressions, which include a convergence term, the equilibrium growth rate is exogenous and is determined by exogenous technological progress and population growth. In neoclassical growth models, the savings rate only determines the level of per capita income in the steady state. Endogenous growth models, which can, in principle, generate a structural relationship between the long-run growth rate and some aspect of financial intermediation (e.g. the efficiency of financial intermediation), on the other hand, have no room for convergence terms, which are always found to be significant in cross-country growth regressions. Thus, they do not sit comfortably with the empirical findings on finance and growth.
Our main hypothesis here is that movements in private credit and liquid liabilities have more to do with the financial cycle. Financial variables are bound to be correlated with average growth rates, more so during the upswing when a positive relationship between growth and private credit or liquid liabilities is to be expected. Growth in private credit fuels consumption and asset prices, raising collateral values and, through their wealth effects and relaxation of credit constraints, fuels consumption further. In empirical studies, this will show up as a positive association between average GDP growth rates, private credit and liquid liabilities but such an association is spurious, although the financial cycle upswing may be long-lasting. During the financial cycle downswing, this association weakens, as repair of balance sheets through deleveraging is a lengthy process, while, during a balance sheet recession, growth declines much more rapidly. During the downswing, it is also possible to observe a negative relationship between private credit and growth due to non-synchronous dynamics: private credit can start declining, while the economy is beginning to recover, if external demand or expansionary fiscal policy outweighs the reduction in firm investment and household consumption. Such a reversal is, in fact, present in more recent studies of the finance-growth nexus. Demetriades et al. (2017), for example, provide new evidence on the nexus showing that private credit to GDP enters with a negative and highly significant coefficient in cross-country growth regressions for the period 2000-2011.
Figures 1, 2 and 3 illustrate some of the above points diagrammatically, in that they provide scatter plots of the finance-growth nexus during 1960-1979, 1980-1999 and 2000-2015, respectively. These graphs depict average growth rates and private credit to GDP for al countries that are available in World Development Indicators; they are, therefore, an unbalanced panel of countries as the number of countries has grown over the period. Nevertheless and perhaps surprisingly, the scatter plots show a clear flattening of the finance-growth nexus over time. Further evidence of the reversal of the relationship between private credit and growth rates is shown in Table 1, where we also present correlations for a balanced panel of 98 countries.
While the 2000s reversal to some extent appears to reflect the entry of new countries into the sample, it remains negative even when the balanced panel is used, probably reflecting the crises that occurred in the 2000s in countries with large financial sectors. Whichever sample is used, it is clear that the finance-growth nexus was only positive and strong in the 1960s and that it weakened considerably in the 1970s, recovered somewhat in the 1980s, weakened again in the 1990s and was reversed in the 2000s.

Capture by Ruling Elites
The standard political economy view of the relationship between finance and growth is that financial development is thwarted by ruling elites: financial and industrial incumbents stand to lose from greater competition that financial development will, somehow, foster (Rajan and Zingales 2003). This view, known as the Rajan and Zingales hypothesis, may well go some way in explaining the lack of financial development in some parts of the world, although the empirical evidence suggests that it does not hold universally. Specifically, Baltagi et al. (2009) provide evidence from 42 developing countries that is only partially supportive of the Rajan and Zingales hypothesis.
In sub-Saharan Africa, for example, what appear to be holding back financial development are acute institutional constraints and tensions along ethnic lines (e.g. Andrianova et al. 2017). Moreover, the Rajan and Zingales hypothesis does not fully account for countries in which financial development actually took off, at least as reflected in the standard variables that have been used to measure it, i.e. private credit and liquid liabilities. This has been the case in many middle-income countries in the last three decades but, also, in many developed economies where the expansion in bank leverage is now thought to have been a contributing factor to accumulation of systemic risk that led to the global financial crisis.
Why do financial and industrial incumbents allow this to take place? Could it be because they actually benefit from such growth? Indeed, our conjecture here is that growth in bank liabilities and private credit need not threaten financial and industrial incumbents. The growth in bank liabilities and private credit can take place without increased banking or industrial competition. If it does, the 'party' for the ruling elite is an even bigger one. There are more 'soft' loans for incumbent industrialists and real estate developers who can then continue building bigger empires. There are more profits and bonuses for bankers by lending to households to buy the products of the industrialists and the homes built by developers. It is a merry go-round, which increases consumption and business expenditure, including construction, although not necessarily productive investment, particularly if industrial incumbents manage to fend off competition from new entrants. Such a 'party' is unlikely to continue forever even in the absence of competition, not least because it is likely to create property price bubbles and over-indebtedness, all of which will sooner or later be reflected in growing NPLs on bank balance sheets. The inefficiencies inherent in uncompetitive markets and the rent seeking by incumbents are likely to prevent productivity growth from lifting the economy out of trouble. If ruling elites benefit from larger bank balance sheets and lending booms, then it is not surprising that the larger international banks, which stand to benefit most from 'too big to fail' policies (that artificially reduce the cost of raising liquid liabilities), would exert undue influence on financial regulation. Seen in this light, Basel II, which effectively allowed large international banks to set their own capital requirements-by letting them use their own risk models to calculate their riskweighted assets-seems easier to understand. We conjecture that Basel II was the outcome of a process that favoured the big banks and allowed them to take on excessive risks that eventually led to the global financial crisis. 7 Against this context, the findings of Andrianova et al. (2008Andrianova et al. ( , 2012 relating to the robust positive association between government ownership of banks, financial development and growth should be of little surprise. While anecdotal evidence suggests that bureaucrats are generally bad bankers, the authors' findings reveal that such anecdotal evidence cannot be generalized. Conditioning on the usual determinants of long-run growth, Andrianova et al. (2012) find that countries with government-owned banks have, on average, grown faster during 1995-2007 than countries with little or no government ownership. The authors suggest that this may be because corrupt politicians in democracies might find it easier to extract rents from poorly regulated private banks than from government-owned ones. Stateowned banks, in addition to facing the same regulation as private banks, are either directly or indirectly subjected to parliamentary scrutiny. Often, they are also subject to public-sector audits. They and can, therefore, be compelled to avoid excessive risk taking, as well as being prevented from making political contributions. It may, therefore, be in the interest of rent-seeking politicians to opt for a regime of weak prudential regulation and supervision, which, by failing to address agency problems in private banks, could allow them to extract rents.

Institutional Weaknesses
Institutional weakness is perhaps most acute in sub-Saharan Africa, where the finance-growth link is largely missing (e.g. Demetriades and James 2011). In such circumstances, financial deepening, by itself, is unlikely to result in more growth. Andrianova et al. (2015a, b) show that low institutional quality is, in fact, the main contributory factor explaining why African banks lend so little, which in turn provides a plausible explanation for banking sector underdevelopment in SSA. Specifically, they put forward a theoretical model in which a bad credit market equilibrium arises because of the combination of asymmetric information problems with low institutional quality relating to loan contract enforcement. Andrianova et al. (2015a, b) provide supporting evidence of their theoretical results by applying a dynamic panel estimator to a large sample of SSA banks. Specifically, their 7 Evidence that larger banking systems are more prone to crises is provided in recent research by Demetriades et al. (2017), which shows that the ratio of private credit to GDP is among the strongest determinants of the probability of a systemic crisis in a sample of 121 countries over 1999-2011. findings reveal that loan defaults inhibit bank lending but only when institutional quality is below a certain threshold.
Earlier work has, in fact, shown that the effects of finance on growth are dependent on institutional development. For example, Demetriades and Law (2006), using data from 72 countries for the period 1978-2000, find that financial development is able to promote economic growth only when the financial system is embedded within a sound institutional framework. They also show that in LICs, where institutional quality is low, the relationship between finance and growth is at its weakest. In a similar vein, Deakin et al. (2010) show that strengthening creditor rights in India during the 1990s and 2000s led to an increase in bank credit, supporting the view that legal system improvements can help shape financial development.
Institutional weaknesses are not, however, confined to the poorest countries in the world. Even in developed economies, institutions that are essential for the smooth functioning of credit markets such as foreclosure and bankruptcy laws and banking supervision can be highly imperfect. The recent crises in Europe, including Ireland, Spain, Greece and Cyprus, are very indicative of such weaknesses. Greece and Cyprus, in particular, are the two countries in Europe with the highest ratios of nonperforming loans, largely because their foreclosure laws have traditionally offered a very high degree of protection to borrowers. Lenders can take ten or more years before they can foreclose on defaulting loans, which is a practice that encourages strategic default among borrowers. Legislators have been reluctant to reduce that protection, notwithstanding EU/IMF conditionality attached to their respective recent bailouts, not least because of political constraints, if not political economy considerations (e.g. the protection of real estate developers).
A corollary of the above analysis is that bank liabilities and private credit are too narrow a measure of genuine financial development. We certainly need broader measures that capture the efficiency and soundness of financial intermediation. The measures we have been using in the last three decades have, in effect, been more closely associated with the build up of systemic risk and financial instability than with genuine financial development.
Institutions can, of course, hardly be considered exogenous even in the medium run and are likely to reflect the preferences of the ruling elite. The case study to which we now turn provides ample evidence to this effect.

Case Study: The Political Economy of the Cyprus Banking Crisis 8
A good example of a ruling elite benefiting from rapid 'financial development' and shaping institutions accordingly is the case of Cyprus. During 2005-2011, the Cypriot banking system's assets more than doubled in size, reaching €141.2 billion largely due to the influx of foreign capital. Relative to the size of the economy, total banking sector assets peaked in 2010 at 9.5 times GDP. Domestic banking sector assets, which are more representative of the contingent liability for the taxpayer, peaked at over 6 times GDP and, as such, were the highest in the European Union. Two-thirds of these assets were accounted for by the two largest banks-Bank of Cyprus and Laiki. The former had a balance sheet size of 2.1 times GDP, while the latter's size was 1.9 times GDP; relative to GDP, they were the two biggest banks not just in Cyprus but in the entire EU. Not surprisingly, perhaps, these two banks and the power elite closely connected to them were able to exert considerable influence on the political process and the media, which shielded them from effective scrutiny by the regulators.
During 2005-2012, bank credit to the private sector-using the standard WDI indicator of financial development-increased from 154.4 to 249.6% of GDP. By 2012, Cyprus had the highest ratio of private credit to GDP worldwide. 9 It was followed, although by some distance, by Hong Kong (198.5% of GDP), Denmark (182.0%), USA (179.0%), Japan (175.0%), Switzerland (167.0%), UK (164.5%), Spain (156.8%) and Portugal (152.9%). This comparison would suggest that in 2012, Cyprus was the most financially developed country in the world. The reality, however, was very different. The flip side of the highest ratio of private credit to GDP was an over-indebted private sector and deteriorating asset quality in bank balance sheets. By December 2011, private-sector indebtedness had reached 286% of GDP and was the third highest in the EU.
The banking crisis, which erupted in mid-2012, was triggered by losses from the two largest banks' investments in Greek Government Bonds (GGBs), as a result of the Greek debt write-down that took place in late 2011. Contrary to popular myth, however, this was not the cause of the crisis. The investments in GGB's, as explained in Demetriades (2017a, b), were, in effect, a 'gamble for resurrection', as the banks' interest income from their lending operations dwindled due to rapidly growing NPLs, which peaked at over 50% of total loans. Such record NPL ratios were largely the result of imprudent lending practices and were exacerbated by the bursting of the real estate bubble in 2008-2009, following the onset of the global financial crisis. Lax lending standards, which contributed to the formation of the property bubble in the first place, reflected lax corporate governance within banks, which, in turn, was symptomatic of weaknesses in prudential regulation and supervision. Combined with the abundant liquidity generated by capital inflows, this created fertile ground for excessively risky lending resulting in financial fragility that eventually led to the crisis.
The large volume of capital inflows during 2005-2011, which could be misinterpreted as rapid financial development, and the institutional set-up for prudential regulation and supervision had their root causes in the 'business model' adopted by the ruling elite of the country (Demetriades 2017a, b), which specifically targeted capital flows from the Russian Federation and other former Soviet Republics. The central role in this model was played by politically connected law firms that acted as 'introducers' of wealthy clients from former Soviet Republics looking for a tax-efficient safe heaven for their wealth. The business model included the banks, especially the larger ones, law and accountancy firms and politicians; many leading law firms were, in fact, politically connected, often to successive governments or parliament. The wider business community, including real estate developers, although not part of the 'business model', benefited through access to easy credit. Easy credit to households facilitated home ownership as well as the acquisition of holiday homes and luxury cars and other consumer durables. Capital inflows helped to further fuel aggregate demand and the property bubble. This unprecedented upswing in the financial cycle was accompanied by large increases in GDP and a seemingly healthy fiscal balance, which reflected increased government revenue receipts from indirect taxation and taxes on property sales. There was, however, a growing current account deficit, reflecting a deteriorating trade balance due to excessive consumption of imported goods and services. Nonetheless, the ruling elite congratulated itself for the 'economic miracle' it had created and the finance minister in charge of the economy during this unsustainable boom was credited with the (artificial) improvement in public finances.
A superficial analysis of the positive association between private credit and GDP that emerged during such an unsustainable boom is consistent with the spurious relationship between finance and growth that we alluded to in the previous section. At the same time, such 'financial development' was the flip side of rising, albeit hidden, financial fragility since banking risks were underestimated or altogether ignored [in line with Minsky's (1992) financial instability hypothesis]. There were, nevertheless, some warnings but these were readily dismissed. The two big banks had marketing and advertising budgets that were commensurate to their grossly inflated balance sheets. Because of that, they were able to exert considerable influence over the media. Media owners showed no appetite to criticize banks; critical journalists were silenced or dismissed. Media capture went hand in hand with the capture of nearly the entire political system; at any rate, key parts of the political system were central in the business model. (The politically connected law firms were, in fact, responsible for the influx of foreign capital.) Financial regulation was and remained lax; for example, the definition of NPLs and loan loss provisioning fell well below international standards; independent directors on bank boards were a small minority. The authority responsible for banking regulation and supervision was the national central bank. Although independent, the Central Bank of Cyprus (CBC) was accountable to the national parliament. Interestingly, however, the CBC was hardly, if ever, scrutinized in parliament for its lax supervision of excessive risk taking by banks, even when the credit rating agencies started downgrading the sovereign for the large contingent liability represented by overblown balance sheets. That regulation was and remained lax was no coincidence: it was an endogenous equilibrium that suited the needs of the ruling elite; as such it was not disturbed. Ex post facto, however, the ruling elite blamed the CBC for its lax supervision and took actions to reduce the central bank's independence. Such populist actions appeared to be justified, although they failed to acknowledge that the root cause of the problem was not so much excessive independence but inadequate accountability, for which parliament (in which the interest group behind the business model was more than adequately represented) was solely responsible.
When the crisis hit, it took the ruling elite by surprise: 'financial development' and the 'economic miracle', which emanated from foreign inflows, turned into financial fragility and a full-blown banking crisis within a space of a few months. Deposit flight started from Laiki in the autumn of 2011, and the bank resorted to the Eurosystem for emergency liquidity assistance (ELA). By June 2012, when the two big banks failed to address their capital shortfalls, Cyprus was forced to apply for an IMF/EU economic adjustment programme, not least because the ECB could otherwise cease to continue supplying liquidity to the banking system. However, it took several months to negotiate the bailout agreement with the outgoing left-wing government refusing to accept labour market reforms and privatization of stateowned enterprises (mainly utilities and the port authority). The election of the new right-wing government in early 2013 created the expectation that the banks would be bailed out at the expense of the taxpayer (whom no political party appeared keen to represent). The bankers themselves expected to be spared of the consequences of their actions, as the new government was much closer to the 'business model' than was the case with the previous government. The first attempt to spread the burden of saving the big banks to small savers-including insured depositors-was initially made by the new government. When confronted with the IMF's debt sustainability analysis, which showed that the two big banks were simply too big to save, the newly elected government proposed a 9.9% levy on uninsured deposits and a 6.75% levy on insured deposits to generate the funds to shore up the capital buffers of the two biggest banks. This was presented as an attempt to spread the burden of saving the banks evenly, but it was, in essence, a disguised attempt to protect the 'business model' by limiting the impact on foreign (mainly Russian) depositors. This attempt failed as the proposed levy was turned down by parliament, although perhaps for all the wrong reasons. When the government realized this 'deposit haircut' was politically damaging, it distanced itself from it by claiming that it was forced to accept it by the ECB, which allegedly threatened to cut off liquidity from Cypriot banks. As a result, the proposal to tax deposits was perceived-erroneously-as the result of an ECB 'blackmail' and was turned down on the expectation that Europe and the IMF would somehow change their mind and accept a more traditional bailout with the burden of saving the banks falling on taxpayers. However, as that alternative had already been considered and rejected by the IMF on grounds of public debt unsustainability, the only remaining solution was to bail in junior creditors and uninsured depositors at the two banks. This was eventually accepted by the government, for the only alternative was sovereign bankruptcy and exit from the Eurozone. Nonetheless, the ruling elite proceeded to create scapegoats and launched an unprecedented campaign of misinformation against the CBC, which by then had been given resolution powers, which included changes in the legal framework of the CBC the eroded its independence. These changes eventually led to the resignation of the CBC Governor and resulted in further erosion of the independence of perhaps the only institution that was not directly under the government's influence.
This case vividly demonstrates that a ruling elite may well benefit from rapid expansion of private credit, which may be perceived as 'financial development'. Such 'financial development' may, temporarily, at least promote economic growth, through the financial cycle upswing. It also demonstrates that financial development of this kind can easily turn into financial fragility and a full-blown banking crisis, when as a result of such rapid expansions in bank balance sheets, banks become 'too big to regulate', not least because of the support they receive from the ruling elite. It finally demonstrates that institutions such as financial regulation and rule of law are highly endogenous and can easily be adapted to serve the interests of the ruling elite.

Concluding Remarks
The early empirical literature on the finance-growth nexus is now known to have overstated the possible effects of financial deepening on economic growth. Often the causality issue was downplayed, while authors warning against the use of crosscountry regressions to establish causality were brushed aside or ignored. The research was, nevertheless, influential in terms of creating a near consensus among international financial institutions that helped to promote policies of financial deregulation and liberalization, as they were seen as necessary to promote growth through financial deepening. 10 In developing countries, banking systems were liberalized by removing controls on interest rates, freeing capital flows and privatizing government-owned banks. In developed economies, a relatively laissezfaire stance towards financial regulation was adopted, which, among other largely negative consequences, effectively allowed large internationally active banks to set their own capital requirements. These developments resulted in financial fragility and crises in many countries but failed to promote genuine financial development in the poorest nations in the world.
Notwithstanding the experience of the last two decades, we have no doubt that well-regulated banking systems embedded in appropriate institutional frameworks, including respect for the rule of law, effective banking supervision and good corporate governance in banks, can play a significant role in promoting growth and poverty alleviation. Our analysis suggests that for banking systems to work as well as is intended in theory, the influence of ruling elites on institutions including the rule of law and financial regulation and supervision needs to be minimized. Since these influences are often well hidden from public view, policies that enhance transparency in public life and financial education should be strongly encouraged. The economics profession itself, which has been widely criticized for failing to predict the global financial crisis, should take the lead; otherwise, there is the risk that it will be perceived as the protector of ruling elites (and consequently marginalized). Specifically, the profession-including its 'leading' academic journals-needs to genuinely reform itself to become-and to be seen as-more transparent and less susceptible to influence from powerful interest groups.