---

Macroeconomic Policy

Index

• Estimating SARB's Policy Reaction Rule
• Identifying aggregate supply and demand shocks in South Africa
• On the Rand - Determinants of the South African exchange rate
• South Africa - Macroeconomic challenges after a decade of success
• The Cyclicality of monetary and fiscal Policy in South Africa since 1994
• Through the pass-through - measuring central bank credibility

---

Estimating SARB's Policy Reaction Rule

The paper makes the point that a key problem of monetary policy is one of finding a credible anchor that does not jeopardise the central bank’s ability to react to shocks. It also points out that central banks tend to shy away from combining targets (like inflation AND the exchange rate), something that may have to do with the credibility loss associated with providing a weak signal on the intentions and instruments of monetary policy.

The central question of the paper is whether the SARB’s actions respond to its repeated claim that it does not care about movements in the exchange rate.

The paper points out that, as with other monetary policy anchors, inflation targeting can be ambiguous, firstly because of a lack of full agreement on the main conditions and features of inflation targeting and how they apply during transition to low inflation, and secondly, because some countries have used simultaneously inflation targets and other nominal anchors. In addition, inflation targeting is in practice a broad category that includes a large array of alternative varieties.

It is an inescapable fact that monetary policy under inflation targeting cannot neglect exchange rate fluctuations, but there are real identification problems to distinguish between foreign exchange intervention and exchange rate targeting.

The main question that the paper then sets out to address, is therefore how the presence of the exchange rate can be identified in the reaction function of the SARB.

The paper uses a Dynamic Stochastic General Equilibrium (DSGE) model for South Africa, and compares this with the estimates for other countries from related work. The model is described in detail in an appendix to the paper.

Based on the analysis, the paper finds the results to be consistent with the explicit views of the SARB, i.e. a staunchly anti-inflation bias and low preference for exchange rate fluctuations. A word of caution is raised in that the model, as stated, captures the changes in the policy instrument of the SARB in response to inflation, output and exchange rate dynamics. This means that the model will miss interventions geared to control the exchange rate that do not occur through this channel, e.g. the use of capital controls or interventions in the forward market, both practices which have been common in South Africa.

The model’s comparison with other countries shows that the SARB appears to be on the low side in terms of its concern for the exchange rate. This would be consistent with South Africa having avoided the “original sin” that would preclude it from issuing debt in its own currency. As a result the SARB appears to be distinctly inattentive to what happens with its exchange rate, compared to other developing countries.

An analysis of the stability of the reaction function across time also suggests that the SARB has been able to build a tradition of a stable policy reaction function. In particular its anti-inflation bias has been among the steadiest.

In summary, the paper estimated the policy reaction function of the SARB. It found monetary policy to be quite similar to that of Canada and the UK, and close to that of Australia and New Zealand. Relative to other emerging countries, it stands out for its stability and its relative stronger weight on output and lower relative weight on the exchange rate. It also shows a strong anti-inflation bias that appears to be among the steadiest among emerging economies.

---

Identifying aggregate supply and demand shocks in South Africa

The paper studies output fluctuations over a period (1960-2006; 1983-2006) and then decomposes these output fluctuations into aggregate demand and supply shocks over the same period. It then uses the aggregate demand shocks as a new measure of the business cycle, and the aggregate supply shocks as a novel estimate of potential output.

The paper aims to contribute to the SA government’s ongoing attempts to identify constraints to economic growth with the goal of raising the sustainable rate of growth of the economy and employment creation, by estimating the current potential GDP growth rate.

The paper provides a literature overview, followed by an exposition of the methodology to identify the various output shocks. Thirdly it describes the data used, and then discusses the empirical results.

Literature

With one exception (2984) empirical literature on potential GDP in South Africa is fairly recent. A variety of methods is used in the literature. However various methods yield strikingly similar estimates of potential GDP growth.

Methodology (SVAR analysis)

The paper uses a modified version of a model developed by Blanchard and Quah, with three variables, i.e. real GDP, and two demand shocks, one interpreted as a fiscal policy shock, and the other as a monetary policy shock. This section also sets out the technical assumptions as to the impact of fiscal and monetary shocks on real GDP, and of monetary shocks (the real interest rate) on government consumption relative to GDP.

Data

Two data series are used, one covering 1960-2006, and the other covering 1983-2006. The data selected for the fiscal policy stance is government consumption and not the deficit. The real interest rate is used as a proxy for monetary policy.

Results

The data show that actual supply shock responses match the theoretical model. Supply shocks have a permanent or long run impact on real GDP while the two components of aggregate demand (fiscal and monetary demand shocks) have only transitory impacts.

Two deductions are made from the Data. Firstly, the long-run development of real GDP is dominated by the history of supply shocks. Secondly, over the short to medium term, fiscal shocks dominate monetary shocks in their impact on real GDP, but only in the model estimated on the longer sample (1960-2006). In the post-1983 sample the relative importance of monetary and fiscal shocks are reversed. The pivotal point occurs around 1980 in the monetary history of South Africa. The paper connects this change with the reforms of the monetary policy regime by the De Kock commission and implemented in the 1980s.

A positive supply shock raises the real interest rate temporarily. A positive fiscal shock lowers the real interest rate temporarily.

An interesting emerging fact is that in the period between the mid-seventies and mid-nineties, potential GDP growth declined by a cumulative 30% - a unique quantification of the costs of maintaining apartheid since the 70s, in a time of adverse international shocks. Once again the data analysis gives a plausible pattern and magnitude, measured against other historical accounts.

After 1994 potential GDP growth started rising almost immediately. Both datasets show potential GDP growing by a healthy rate after the political transition, although well short of the goals set by the SA government.

An important conclusion drawn from the data is that the potential GDP growth is between 2.5% and 4.4% while 3.5% would be a likely point estimate.

Conclusion

Two important conclusions are drawn from the data analysis, by the paper. The first is that the cost to the economy of the apartheid policies from the seventies to the nineties was a cumulative 30% decline in potential GDP growth. The second conclusion is that at the time of publication of the paper, the potential GDP growth of the SA economy is around 3.5% per year. This is well below the growth rates envisioned by the SA government.

---

On the Rand - Determinants of the South African exchange rate

The paper tries to answer some questions about the determinants of the rand’s exchange rate so as to develop a predictive model for rand exchange rate movements. Is the rand a commodity currency, i.e., a currency that appreciates when prices of the mineral products it produces are strong on world markets, and depreciated when they are weak? Or does the rand behave like the currencies of industrialised countries, in light of its developed financial markets?

Did the removal of capital controls in 1995 cause a structural break in the determinants of the real exchange rate?

Has there been an element of momentum to some recent movements, or can they be explained by fundamentals?

The paper finds there has been a relationship between the real mineral price and the real exchange rate; however, the real exchange rate also correlates with relative real output, which means it is difficult to distinguish the effect of mineral prices from the effect of real output, because those two factors are themselves highly correlated.

The model finds a high correlation between real income and mineral prices, and also supports the notion that the rand behaves like an industrialised currency with sophisticated financial markets. However this is because the interest rate differential between South Africa and other currencies attracts speculative inflows, but is an indication of risk (inflation, depreciation and default) rather than of high expected returns inherent in the productive sector.

As for structural changes, although there have been significant changes due to the political transition, notably the capital account liberation and the abolition of the dual exchange rate, the paper finds that the data series is not sufficient to be useful.

Using the model it is found that the actual rand movements matched the model predictions quite closely. This would suggest that the rand was not undervalued in 2006, as was suggested by some.

---

South Africa - Macroeconomic challenges after a decade of success

Introduction

Halfway through the new decade, the South African economy has done very well. Growth was high in 2005, capital inflows and the rand are strong, the budget is relatively healthy, and inflation rates and interest rates are low. As democracy continues to consolidate, there are plenty of grounds for optimism; in fact business confidence indicators and private investment are at an all-time high. This paper asks the question whether such achievements provide grounds for complacency. In other words, is the job done? Or do these achievements open the door to new challenges? Are there risks in the horizon? And how does the government’s ASGI-SA strategy deal with the challenges?

The paper provides four areas of analysis: an analysis of the current account, the consistency of the ASGI-SA programme, the benefits of the current fiscal-macro policy mix and the choice of exchange rate regime.

Current account vulnerability

In contrast to many other emerging economies South Africa’s current expansion has come hand in hand with a large current account deficit, which in the first quarter of 2006 topped 6% of GDP. An optimistic view would hold that the deficit reflects rational adjustment to a new equilibrium of permanently higher commodity prices, high investment, or previously pent-up consumption by a new middle class. A pessimistic view is that the boom is unsustainable, perhaps because of a temporary spike in global mineral prices in 2006, because consumers do not fully understand the restrictions they face, or because there is always the risk of a sudden stop of capital inflows. According to this view it is important to keep imbalances in check and to think about managing the inflows in such a way as to minimize the likelihood and severity of the sort of crisis that afflicted other emerging markets in the 1990s.

Even if the terms of trade have not risen spectacularly – the big rise in prices for South African mineral exports having been substantially offset by a big rise in the price of oil imports – the global commodity boom, together with the emerging market boom, is nonetheless largely responsible for the appreciation of the rand. The reason is that investors have piled into South African assets (especially equities), thus bidding up their price, not only in the form of higher rand prices of equities but also in the form of an appreciation of the currency. Easy money emanating from the world’s major central banks (Fed, BoJ, ECB, and PBoC) over the period 2001-2005, together with a possible bubble component over the period 2005-06, have probably been one force behind the movement into commodities generally, emerging markets at large, and commodity based emerging markets in particular.

The real appreciation of the rand is in turn one reason for the country’s large current account deficit. The bad news is that the bubble component may be especially applicable to South Africa because most other emerging markets are running trade surpluses now. The good news is that even if a crash comes – spring 2006 saw a hint of it -- South Africa is well-positioned to weather it for a number of reasons: borrowing has been a relatively small fraction of the inflow (versus equity and FDI), the rand floats (though not without substantial intervention), and much of the debt is rand-denominated. These factors offer good grounds for hope that a correction will be fairly automatic and not involve the painful “currency crash plus balance sheet contraction” so familiar from the emerging market crises suffered by other countries in 1994-2001. While a large crisis is off the books, the paper provides a series of simulations that suggest that a significant reversal in flows would nevertheless impact the South African economy by slowing growth, hurting investment and, potentially, worsening fiscal accounts. In addition it is shown that the current growth scenario requires increasing terms of trade just to keep the current account in recent levels. If the terms of trade were to stabilise – let alone deteriorate - the current growth dynamics would lead to a significant deterioration of external accounts.

Thus, the paper can be read as sending a cautionary note on the need to reduce the external imbalances of the economy. It also provides some policy recommendations to further minimize the negative impact of a possible sudden stop. For example with regard to debt management the advice is that South Africa seek further to reduce the share of capital inflow that takes the form of short-term and dollar-denominated debt, by taking advantage of its ability to borrow in rand and to attract equity and FDI inflows. As a more novel proposal, the suggestion is that, when the South African government undertakes major borrowing, it may want to consider denominating some of the liabilities in terms of gold, platinum, or prices of other major mineral exports, in hopes of getting such a market going.

The consistency of ASGI-SA programme

To make a statement on the “macro-consistency” of ASGI-SA the paper reviews successful growth transitions. In particular, Rodrik (1998) shows that successful growth accelerations have the feature that for each 3% percentage points of increase in output growth, investment to GDP ratios increased by just 1%. This implies that growth has come much more from productivity improvements than capital deepening. However when analysing South Africa’s growth programme it is found that these numbers are inverted: a 1% increase in the growth rate requires a 3% increase in the investment to GDP ratio. This is not entirely unexpected. Given the employment/productivity performance of the South African economy even such large investment programme will barely deliver the desired growth rates while imposing an impossible burden on public investment. All these are important problems even before discussing the feasibility of increasing domestic savings to finance this programme without increasing further the external vulnerability of the economy. As a result the paper argues that the ASGISA programme has serious macro inconsistencies that need to be discussed.

The fiscal-monetary policy mix

The paper argues that the point of coordinating fiscal and monetary policy is primarily to introduce countercyclical damping to the business cycle because this brings growth gains over the longer term.

An evaluation of the business cycle features of fiscal and monetary policy finds that so far fiscal policy has been mostly pro-cyclical, whereas monetary policy has shown less evidence, over the last couple of years, of the mild countercyclical pattern it had established during the late nineties. The central bank appears to be increasingly enthusiastic about its successful inflation targeting strategy. The paper shows, however, that when an economy faces important supply shocks inflation targeting exacerbates the business cycle and not the other way around. Inflation targeting is fine, but any increases in the CPI attributable to increases in dollar prices of imports, or negative supply shocks, should not prompt monetary tightening to prevent prices from rising. For example, a recent sudden stop, by depreciating the Rand, has rekindled inflation fears at the central bank prompting it to raise its discount rate. This is considered to be not an optimal policy response. An analysis of South Africa’s output response to fiscal and monetary shocks shows that monetary policy is the most convenient tool for implementing countercyclical macro policy. The central bank has earned the credibility to push in this direction. In contrast fiscal policy should remain “passively” countercyclical -a counter cyclicality that should arise from a relatively stable expenditure pattern combined with pro-cyclical tax revenues. The paper argues that such countercyclical fiscal policy provides growth gains.

Exchange rate regimes

The exchange rate regime choice is intimately related to the objectives of monetary policy. The paper recommends that the authorities continue to float, so that they can pursue a countercyclical macro policy, but also continue to manage the float somewhat by intervening to stabilize the exchange rate. Cross country evidence suggests that fighting exchange rate appreciation is associated with higher growth in the future. As a result the paper proposes an active exchange rate policy to keep the currency from becoming overvalued. This may include some consideration of capital controls that are aimed at controlling inflows to mitigate the risks of a “sudden stop” scenario, while at the same time removing any remaining barriers to outflows.

---

The Cyclicality of monetary and fiscal Policy in South Africa since 1994

This paper uses a structural vector auto regression (SVAR) approach to discuss the cyclicality of fiscal and monetary policy in South Africa since 1994.

It starts with a literature study of the topic. The first decade of democratic South Africa has seen the longest business cycle expansion in the country’s history. This “great moderation” was characterised by lower and stable inflation, lower and stable real interest rates, positive and steady GDP growth, and stable fiscal deficits and debt. This period of moderation compared well with other developing economies and not only against South Africa’s own economic history. One of the factors associated with the period of moderation is the fact that the economy has become progressively more open to international capital flows and trade. Openness results, firstly, in a reduction in the effectiveness of monetary policy because prices become linked to the exchange rate and there is a reduced need to pursue inflationary policies. Secondly, openness allows for quicker recoveries from external shocks. Finally, openness reduces the possibility of a “sudden stop” for smaller countries. Another factor associated with moderation is the successful political transition.

Though not undisputed, there is growing consensus that monetary policy has contributed to the remarkable stabilisation of the South African economy over this period. The evaluation of the role of fiscal policy in stabilisation has been less favourable and there is little evidence that a countercyclical fiscal stance was a priority over this period. This paper considers these issues in an empirical framework that addresses some of the shortcomings in the literature. Specifically, it constructs a structural model in contrast with the reduced form models typically used in the South African literature, incorporates the dynamic interaction between monetary and fiscal shocks on the demand side and supply shocks on the other, and avoids controversy over ‘neutral’ base years and the size of fiscal elasticities. The model confirms the consensus on monetary policy, finding it to have been largely countercyclical since 1994. On fiscal policy, this paper finds evidence of pro-cyclicality, especially in the more recent period, though the policy simulations suggest that the pro-cyclicality of fiscal policy has had little destabilising impact on real output.

“The stakes in choosing the right monetary and fiscal policies are high” Robert Lucas (2003: 3) argued in his presidential address to the American Economic Association. Mistakes at this level cause inflation, distort decisions in labour and capital markets, and might either precipitate or fail to prevent recessions. Lucas proceeded to calculate the potential welfare gain of improving fiscal and monetary policies to attain an ‘optimal’ stabilisation policy and found that the potential welfare gains from such improvements were orders of magnitude smaller than the potential gains from supply-side policy reforms. Whether this result and the associated policy priorities also hold for South Africa is an empirical question, and this paper contributes towards an answer thereto by estimating (i) the extent to which monetary and fiscal policies have been pro- or anticyclical since 1994 and (ii) the effect of these cyclical characteristics on output volatility in South Africa.

This paper examines the cyclicality of monetary and fiscal policies in an empirical framework that distinguishes between transitory ‘aggregate demand’ and permanent ‘aggregate supply’ shocks, building on the earlier work of Shapiro and Watson (1988), Blanchard and Quah (1989), and Clarida and Gali (1994). Using additional identifying restrictions, the paper disaggregates the demand shock into separate fiscal and monetary policy shocks. These results are described briefly in the second section of the paper.

The third section of the paper uses the identified monetary and fiscal policies to answer the questions about the cyclicality of policy posed above. A counterfactual policy simulation is used to judge the potential gains from more consistent counter-cyclical monetary and fiscal policies in South Africa.

The paper acknowledges that the results from the analysis remain somewhat ambiguous, in line with the literature study. Although it finds that monetary policy has been more anti-cyclical and fiscal policy has been more pro-cyclical, however the model simulation did not predict that the pro-cyclicality of fiscal policy has had a quantitatively large destabilising impact on real output since 1994. The same is found for monetary policy, i.e. it either had no destabilising impact on real output, or has stabilised output over the period since 1994.

---

Through the pass-through - measuring central bank credibility

In open economies, exchange rate fluctuations affect the behaviour of inflation. This makes the exchange rate pass-through (defined as the effect of exchange rate changes on domestic inflation) an important consideration with respect to monetary policy.

The paper sets out to measure the pass-through of exchange rate fluctuations, to inflation, in order to establish the credibility of the central bank’s pursuit of monetary policy. The rationale is that a successful inflation targeting policy would result in a small pass-through of exchange rate fluctuations into the inflation rate. The credibility of the central bank then lies therein that its inflation forecasts are accurate and that exchange rate shocks therefore do not result in interest rate adjustments. A central bank with low credibility would be forced to make interest rate adjustments to compensate for exchange rate shocks.

In the decade between 1997 and 2007 the SA Reserve Bank has been carrying out an aggressive inflation targeting programme. The paper points out that a stable inflation rate is not necessarily the correct measure of how much credibility the central bank has gained, because inflation may be influenced by global movements, and not be the result of internal policy. The paper offers a proposal to measure the pass-through by analysing pricing decisions at firm level, and uses this analysis to argue that over the measurement period there has been a substantial reduction in the pass-through rate and that therefore indeed the SARB has gained significant credibility in recent years (preceding 2007).

The paper considers, but discounts, other possible explanations for the reduction in the pass-though, and concludes that the most likely explanation is the success of monetary policy and therefore the credibility of the monetary authority’s inflation targeting.

The paper concludes by pointing out that more research is required and that the conclusion made is no more than a conjecture.