استقلال بانک مرکزی : نرخ تورم پایین بدون هیچ هزینه؟ یکی از تمرین های شبیه سازی عددی

The independent nature of the Central Bank is often associated with achieving low and stable inflation. Further to that the merits of independence are stretched to achieving low(er) output variability when compared to a government run monetary policy. In this paper we use the Alesina (1989) and Alesina and Gatti (1995) model to examine how often an Independent Central Bank can achieve an improvement on both counts. To do that we run numerical simulations where we change the ex ante probability of elections (and hence the degree of electoral uncertainty) with a view to determining how the private sector’s perceptions affect the level of output variability. Our conclusions agree with the Alesina and Gatti assertion that there will exist occasions when all political parties will be better off by consenting to the running of monetary policy by an independent institution but more often than not this comes at some cost to output. On theoretical grounds therefore, the trade-off between inflation and output variability (à la Rogoff) is still a valid one.

As more and more monetary authorities are granted political and economic independence, the gains in inflation achieved are measured against the ability to sustain and promote growth. And while the benefits in terms of inflation are widely accepted, there is no consensus as to whether output variability also benefits. A number of theoretical attempts find that Central Bank independence2 (CBI) comes at no cost to output variability. Alesina and Summers (1993) find no relation between the variability of output growth and legal independence within industrialised countries and therefore, advocate that independence is an institutional feature worth adopting. Crosby (1998) reverses the causality by arguing that countries that have lower output variability are more likely to choose to have an independent Central Bank. Rogoff (1985) in turn finds a clear trade-off between inflation and stabilisation policy. And since the former benefits from greater independence, the latter suffers. Alesina (1989) and Alesina and Gatti (1995) (from now on AG) justify the output gains of CBI by linking inflation and output variability to electoral uncertainty. They thus argue that in a world in which political parties are polarised in terms of their preferences, the possibility for time inconsistent behaviour is exacerbated. Keen to correctly anticipate the future, the private sector is adversely affected by this uncertainty and therefore, prefers to delegate monetary policy to a conservative body that is not subject to elections. This has two effects: first it guarantees the achievement of monetary objectives and second, it reduces the degree of uncertainty which affects the variability of output. For an appropriate choice of preference parameters, the authors argue, monetary discipline comes at no extra cost to output. And it is therefore, to the advantage of all political parties to co-operatively establish an independent Central Bank with appropriate preferences that will ex ante deliver both lower inflation as well as output variability. Correct though it is, we will argue in this paper that such a result is also incomplete. The authors present their analysis in a two-party system when the probability of each of the parties getting elected is equal to 50%. This is arguably the highest degree of uncertainty (and therefore one which is interesting in its own right) but fails to acknowledge two relevant issues; first, that the probability that any given party will be elected may in fact be something different to 50% and second, and more important, that this probability may vary from one electoral period to an other. The institutional set-up associated with independence on the other hand, is one that cannot change from one electoral cycle to the next to accommodate alterations in electoral preferences. We will therefore argue that if the benefits of Central Bank independence are linked to electoral uncertainty, then it is important to show how they alter with the likelihood of the outcome of elections. The analysis will show that under different levels of political uncertainty (i.e. varying electoral ex ante probabilities), the benefits of Central Bank independence in terms of output variability are much more difficult to discern than previously argued. Section 2 briefly summarises the AG model. Section 3 then demonstrates under which conditions one can talk about “costless” Central Bank independence in the presence of electoral uncertainty. We do that for a range of values for the underlying parameters and use numerical simulations to demonstrate the effects. Section 4 summarises the results and briefly discusses the policy implications.

The literature is agreed on the benefits of independence in terms of inflation but is less so in terms of output variability. The attempt by Alesina and Gatti to link such benefits to electoral uncertainty shows that both are attainable, but then for a very particular level of ex ante probability, namely of 50%. But if this probability is in fact different to this level or it changes, as indeed it will, from one electoral cycle to an other, then a Central Bank with fixed preferences is not guaranteed to achieve price stability at no cost to output. We summarise the results of our analysis as follows: • The extent to which an independent Central Bank is successful in attaining both lower inflation as well as less variable output is directly dependent on the private sector’s expectations on which party will win. • Having shown that, then the trade-off between low inflation and output stabilisation remains a very valid one. • The formation of an independent and ultra-conservative Central Bank will achieve lower inflation but then at the cost of output variability. This is perhaps a relevant issue when we refer to the way that the Central Bank has been recently created in Europe. Its rather clearly defined mandate of price stability, coupled with the need to establish reputation and maintain it, could easily identify it with having more conservative preferences than the average of the country Central Banks and by consequence more conservative than the median voter. If true, then attaining low inflation is in direct conflict with output stabilisation. • The value of k clearly matters. If countries operate close to optimal capacity levels, they may find that they are having to face the conflict between achieving low inflation and stabilising output. If however, the level of distortions is high, then an independent monetary authority may provide for both low prices as well as stable output. The implication is that a Central Bank with fixed preferences is not always capable of achieving low inflation at no cost; it depends crucially on how fluid the preferences of the electorate are (as reflected in those of the political parties that are elected) and the welfare losses resulting from uncertainty associated with electoral outcomes. Since the latter is beyond the scope of policy, it follows that there cannot be a big and sustained discrepancy between the actions of an independent Central Bank and the will of the society, which it ultimately serves.11 Policy-making needs to establish a process of evaluating this and allowing for as much flexibility as warranted by prevailing conditions.