Why does central bank independence vary across countries and over time?

This essay was written as a formative part of my undergraduate studies at the London School of Economics.

This paper assesses the various theories explaining the growth of independent central banks (ICB) across the world and highlights factors that might explain variations in central bank independence (CBI) across polities and over time. I will argue that any account of variations in CBI must place heavy emphasis on the role of ideas and interests to provide complete accounts of CBI development.

There has been an increasing trend towards ICBs over the past century, with the most rapid adoption taking place post-1990. More than 90 percent of polities now have CBI as a core tenet of economic policy (Marcussen, 2005). This surge in CBI is part of the broader trend of neo-liberal consensus in which the health of an economy is measured by three indicators; stable GDP growth, low unemployment and low inflation; the last of which has been emphasized as the most important to achieve desirable outcomes (McNamara, 1998, 2002).

Much of the literature on the growth of ICBs notes the desirability of separating politics from macro-economic stability instruments – namely, interest rate setting (Cukierman et al., 1992, McNamara, 1998, 2002, Bernhard, 1998, Moser, 1999, Marcussen, 2005). This is necessitated by the existence of a time consistency problem; political actors have short-term time horizons which may lead them to manipulate the economy in their short-term interests, at the expense of the polity’s long-term interests (Moser, 1999). These short-term interests supposedly arise because of electoral politics – where governments manipulate the economy to produce a boom before an election and then constrain growth afterward, resulting in the political business cycle – and partisan politics – where parties of the left pursue more expansionary fiscal policy than those on the right – which destabilize the economy through the creation of a boom and bust cycle (McNamara, 2002) and create a credibility crisis (Moser, 1999).

The credibility crisis is a particular concern when considering that the neo-liberal consensus emphasizes price stability above all else. Inflation is a function of people’s expectations of future inflation; if people don’t trust a government to pursue low inflation policies, they will anticipate that – and subsequently act as though – inflation will be high, thus creating high inflation. Faced with problems of high inflation, credibility crises and malfunctioning or non-existent banking sectors, the classical demand-driven model suggests political actors will rationally seek to find the optimal policy solution to their policy failure (Moser, 1999, Marcussen, 2005). Moreover, we will know this policy solution is optimal, because sub-optimal institutional designs would be dismantled were they not effective (Marcussen, 2005). The fact that ICBs have been rapidly adopted around the world might suggest, therefore, that governments have found the optimal policy solution and solved the credibility problem with ICBs; suggesting a link between low inflation, stable growth, and CBI, and that short-term time horizons of politicians are overcome through delegation (Marcussen, 2005).

The problem with demand-driven explanations is that they are highly generic; they suggest a path dependency and trajectory towards a homogenous set of institutions cross-nationally that simply cannot be assumed, placing little emphasis other factors that might have an effect on institutional design and the choices of political actors. Whilst functionalist models seem to suggest isomorphism, the reality is that there is great variance between countries, and even within countries, over time. It is also important to remember that variations between ICBs may be caused because of traditions and norms; even where statutes are identical between two polities, and banks have legal independence, their actual independence depends on many unwritten rules (Cukierman et al., 1992).  In order to explain institutional differences, we must consider the role of ideas and interests in policy making; the trend towards CBI cannot be seen as purely demand-driven, but must include supply-side modeling (Marcussen, 2005).

One way to understand ICB variations is through an assessment of the literature on policy transfer and diffusion. This literature highlights the transfer of policies through observation and learning processes; and its unpredictability. Whilst the introduction of institutional choices can be through coercive transfer from international institutions such as the IMF or EU (Dolowitz and Marsh, 2000, Radaelli, 2000), and can, therefore, be exact isomorphic transfers, it is equally possible for political actors to voluntarily and selectively choose institutional setups from other polities, or indeed to simply take the underlying principles and craft entirely new institutional frameworks in an attempt to fit with the unique cultural, social and economic circumstances of their individual polities (Dolowitz and Marsh, 2000, Simmons and Elkins, 2004). This is one likely explanation for differences in CBI across nations and hints at the role of ideas and interests explicitly in this variation.

The role of policy entrepreneurs (Marcussen, 2005) and epistemic communities (King, 2005) cannot be overlooked in the discussion of cross-national and inter-temporal CBI changes. Institutions do not just appear from nowhere, and delegation to ICBs doesn’t occur by accident. King (2005) describes the role of ideas as a three stage cycle; policy failure, paradigm shift, and emulation. When policies fail – in this instance, when there is a credibility crisis – it is a prime time for policy entrepreneurs to suggest alternatives. Critical junctures (Marcussen, 2005) present such a time to intervene and cause a paradigm shift. Marcussen (2005) argues, from a historical institutionalist perspective, that many of the shifts in CBI have occurred as a result of a critical juncture, like the end of WWI and the collapse of Bretton Woods, or the end of the Cold War, where entrepreneurs such as UK and US central bankers were given the opportunity to spread their ideas to other economies. This plausibly shows the importance of ideas very clearly in the role of inter-temporal variations in CBI.

But, a new idea and a policy failure is not enough, and indeed a policy failure isn’t always necessary – there sometimes only has to be a key decision making figure that is able to be influenced, otherwise there will be a maintenance of the status quo (King, 2005). This can be clearly seen in the case of the UK. The Labour Party granted independence to the Bank of England in 1997 amidst a relatively stable economic environment, with no external pressures for reform. Labour didn’t need to delegate for any economic reason, but instead chose to do so because it was politically attractive; they needed to establish their credentials on the economy after several previous Labour governments had steered the British economy into crisis (King, 2005). Clearly the role of interests also play a part in the degree of CBI as well; where there are no formal economic pressures, informal political pressures may explain some of the cross national variation we see. This example also raises doubts about the historical institutionalist and functionalist  perspectives; whilst clearly they have something to add to the debate around levels of CBI, it appears that they cannot provide truly overarching theories alone – there are several examples that break the path-dependent models due to the roles of ideas and interests.

To further illustrate the role of ideas and interests, take the case of the European Union and the European Central Bank. The members of the European Monetary Union have been trying to establish a Europe-wide, stable economy for decades, testing various institutional designs like the European Monetary Snake, and eventually the single-currency, the Euro. Whilst these clearly show the role of the policy failure and paradigm shift cycle – and thus the role of ideas – it is really the role of interests that explain EMU member states’ inter-temporal variations in CBI. This is rooted in the principles of the EU itself – a way of promoting peace across a war-torn continent through tighter economic integration. The interests of EU member states to change their institutional set up several times were arguably principally driven by their political interests, not their economic ones (McNamara, 1998). This compounds the previous argument – that whilst historical institutionalism and functionalism can explain some aspects driving institutional change, they cannot explain them all.

Of course, sometimes actors actively choose not to create ICBs, because they deem it not to be in their interest – be that personal, or governmental. UK Prime Minister Thatcher was a stalwart against the idea of granting independence to the Bank of England throughout her term of office, despite the fact that the UK was experiencing economic turmoil for much of the 1980s, and that her own Chancellor preferred an ICB. She believed it would show her government to be weak if she delegated monetary policy to the Bank (King, 2005). Bernhard (1998) echoed a similar idea several years before King’s paper, highlighting that political actors (in this case, ministers) will only delegate if it is in their interests; when they face a credible threat of rebellion from their backbenchers or coalition partners, and if the preferences of ministers and backbenchers and coalition partners differ. If no such threat exists, there is no incentive (Bernhard, 1998). This is another example of where ideas and interests matter and have caused institutional variations across nations and across time.

Ultimately, it is difficult to identify any one factor – or even a string of factors in one overarching theory – that can explain the variations between ICBs across countries and across time. This is because politics is an actor-based game, and inevitably, these actors have different ideas and interests. It seems obvious that historical institutionalism and functionalism cannot provide complete accounts of these variations, but neither can a purely ideas- or purely interests- based approach – sometimes there is simply a mixture of them all.