Essay on Globalization and Inflation - 300 word essay apa format.

Rethinking The Current Inflation Target Range In South Africa Lumengo Bonga-Bonga, Ntsakeseni Letitia Lebese

The Journal of Developing Areas, Volume 53, Number 2, Spring 2019, pp. 13-27 (Article)

Published by Tennessee State University College of Business DOI:

For additional information about this article

Access provided by Ebsco Publishing (11 Sep 2018 13:19 GMT)

https://doi.org/10.1353/jda.2019.0018

https://muse.jhu.edu/article/702993

T h e J o u r n a l o f D e v e l o p i n g A r e a s Volume 53 No. 2 Spring 2019

RETHINKING THE CURRENT INFLATION

TARGET RANGE IN SOUTH AFRICA

Lumengo Bonga-Bonga

Ntsakeseni Letitia Lebese

University of Johannesburg, South Africa

ABSTRACT

Critics argue that inflation targeting is not an appropriate monetary policy framework for

developing countries. They maintain that developing countries are more susceptible to the negative

effects of external shocks due to the uncertainty perceived by investors with respect to their

political and economic stability. It is in this line that this paper assesses whether the 3%-6%

inflation target is the optimal inflation target band in South Africa. To determine the optimal level

of inflation target in South Africa, this paper follows the methodology developed by Ball and

Mankiw (2002), which rests on the premise that there is a short run trade-off between inflation and

unemployment. Ball and Mankiw (2002) show that there exists a level of unemployment that is

consistent with stable inflation. The unemployment level that corresponds with a stable inflation is

known as the non-accelerating inflation rate of unemployment (NAIRU). Thus, this paper uses an

expectations-augmented Phillips curve to estimate a time-varying NAIRU for South Africa from

1980 to 2015. We use the headline inflation rate and the official unemployment rate based on the

narrow definition to evaluate the appropriateness of the current inflation target range. Quarterly

data from 1980 to 2015 sourced from Quantec is used to this end. The results of the empirical

analysis indicate that, if South Africa were to put in place an inflation target range based on the

NAIRU, it would have to target an inflation rate that ranges from 1.4 to 11.5 percent. This range is

different to the official inflation target of 3% to 6% adopted by the South African Reserve Bank

(SARB). Furthermore, this paper finds that the Phillips curve is not vertical in South Africa, as

actual inflation does not depend solely on inflation expectations. The policy implication of the

findings of this paper is that the South African Reserve Bank should think about revising its current

inflation target, as it is too narrow for an emerging economy. The current low range of inflation

target could have a negative effect on output and unemployment in the country. This paper

recommends that the SARB should rely on the realities of the South African economy rather than

on external concerns when defining the range of inflation target.

JEL Classifications: E52, C13 Keywords: inflation target, NAIRU, unemployment, South Africa

Corresponding Authors’ Email Addresses: [email protected]

INTRODUCTION

South Africa had double-digit inflation rates from the 1970s to the early 1990s

(Casteleijn 1999; Rossouw & Padayachee, 2008). In 1994, following its re-integration

into the international economy, South Africa was subjected to increased political pressure

to reduce its inflation rate to levels which were commensurate with its trading partners

(Banerjee, Galioni, Levinsohn, McLaren & Woolard 2008; van der Merwe 2004;

Padayachee, 2001; Ricci, 2005). Subsequently, the government introduced inflation

targeting as a monetary policy framework in 2000 (Mboweni 1999) Inflation targeting

policy aimed at stabilising prices in South Africa. A number of studies support the view

14

that the pursuit of price stability has been successful in reducing the high inflation rate in

South Africa (National Labour and Economic Development Institute, 2004 Ricci, 2005).

South Africa managed to achieve single digit inflation rates in the mid-1990s for the first

time since the 1960s, (Casteleijn, 1999). However, while the South African Reserve Bank

(SARB) succeeded in reducing inflation, the government failed to realise the

development objectives as set out in GEAR (National Labour and Economic

Development Institute, 2004). South Africa experienced a continued increase in the

unemployment rate from 1980 to 2010, reaching 27% in 2003 (Department of Economic

Development, 2010). In 2010, South Africa was ranked amongst the top 10 countries in

the world that have high unemployment rates (Department of Economic Development,

2010). According to McCord and Bhorat (2003), South Africa has the highest

unemployment rates compared to other emerging countries in Africa, Asia and Latin

America. The rising unemployment may be attributed to the change in monetary policy as

international empirical evidence (Akerlof, Dickens, Perry, Gordon & Mankiw 1996;

Fortin, 2001; Lundborg & Sacklẻn, 2006) indicates that a change in the inflation rate

from a high level to a low level is associated with perpetual increases in the

unemployment rate. Some experts reinforce this observation by noting that at low

inflation rates, the Phillips curve is negatively sloped, implying a trade-off between

inflation and unemployment (Akerlof et al., 1996; Holden, 2002; Lundborg and Sacklẻn,

2006; Padayachee, 2001). The existence of trade-off means that reductions in the

inflation rate are directly correlated with increases in the unemployment rate (Erceg,

2002). Hence, it is essential for a central bank to establish the correct shape of the Phillips

curve for that particular economy and the consequent dynamics of the

inflation/unemployment trade-off before deciding, on the one hand, on the suitable

monetary policy regime to pursue or, on the other, if inflation targeting is the choice, on

the targeted band.

Critics argue that inflation targeting is not an appropriate monetary policy

framework for developing countries (Kahn, 2008). They maintain that developing

countries are more susceptible to the negative effects of external shocks due to the

uncertainty perceived by investors with respect to their political and economic stability

(Ricci, 2005). Investor uncertainty and the requirement of a flexible foreign exchange

market expose small developing economies to negative external shocks more than it does

big and developed economies (Kahn, 2008; Ricci, 2005). In addition, some scholars

argue that if the central bank makes the pursuit of price stabilisation its sole objective,

economic growth and development are inhibited (Kahn, 2008). However, recent studies

(Arkerlof, Dickens, Perry, Bewley and Blinder, 2000; Ball and Mankiw, 2002; Fortin,

2001; Lundborg and Sacklẻn, 2006; Hsing, 2009; Fortin, 2001), on the relationship

between inflation and unemployment indicate that there is a particular point on a non-

linear Phillips curve where the trade-off is optimised and central banks should set the

target for inflation at this specific point. Since a number of authors suggest that inflation

targeting derives its justification from the theory of natural rate (Bernanke, et al., 1999;

Bernanke, 2003), it is clear that the optimal target for inflation rate should be set by

taking into account the level of natural rate of unemployment, which determines the

shape of the Phillips curve. However, the widespread implementation of low-inflation

targets assumes, without proper verification, the principle of a vertical long-run Phillips

curve, whereby inflation has no long-run effect on unemployment. This paper shows that

15

setting inflation targets simply by relying on the idea of vertical Phillips curve may be

misleading if the actual shape of the Phillips curve is not vertical.

Given the fact that unemployment has maintained a growing trend in South

Africa even after the adoption of the inflation targeting policy, the policy has become

very controversial (Department of Economic Development, 2010; Banerjee, Galioni,

Levinsohn, McLaren & Woolard, 2007; Klasen & Woolard, 2008), this paper aims to

evaluate whether the 3%-6% inflation target band is at a level that optimised the trade-off

between inflation and unemployment. An optimal inflation ban should refrain from

generating excess unemployment. The remainder of this paper is organised as follows: In

Section 2, the literature is reviewed. This is followed by the discussion of methodology in

Section 3. The estimation of results is detailed in section 4 while in section 5 we provide

the conclusion and policy implications of the research.

LITERATURE REVIEW

Theoretical Discussion

The debate on the relationship between inflation and unemployment dates back to 1958

when Phillips (1958) found a nonlinear relationship between unemployment and changes

in wages by conducting empirical tests on data from the United Kingdom for the period

1861 to 1951. Based on the results of his study, Phillips concluded that by accepting

some degree of inflation, central banks could maintain lower rates of unemployment

(Van der Merwe, 2004). Phillips’ conclusion led to major disagreements among

economists on the existence of a relationship between inflation and unemployment and,

consequently, the actions that should be taken to address the trade-off between these

economic indicators (Friedman, 1968; Burger and Marinkov, 2006; Fischer, 1996;

Mankiw, 2001).

Monetary economists argue that the Phillips curve is vertical in the long run

(Friedman, 1968; Fischer, 1996; Michie, 2003). According this group of economists, the

trade-off between inflation and unemployment is temporary and mainly results from the

behaviour of workers as they adjust their wage expectations following an increase in the

unemployment rate (Hodge, 2002; Friedman, 1968). In the long run, following the full

incorporation of inflation expectations in wage negotiations; there is an increase in the

inflation rate and the unemployment rate remains unchanged as nominal wages adjust

towards their real rate (Hodge, 2002). The unemployment rate thus returns to its natural

rate and hence, remains stable in the long run (Friedman, 1968:8 and Michie, 2003). A

vertical Phillips curve, as hypothesised by the monetarists, implies that there is no

permanent relationship between inflation and unemployment. Post-Keynesian economists

contest this tenet of monetary economists. They are of the view that workers base their

wages on adaptive expectations; and that a trade-off between inflation and unemployment

depends on how quickly workers adopt their future inflation expectations following

disturbances in the economy and on the bargaining power of employers during wage

negotiations (Holden, 2002; Michie, 2003; Hodge, 2002 and Michie, 2003). The post-

Keynesian economists maintain that at any unemployment rate, besides the natural rate of

unemployment, there will be a trade-off between inflation and unemployment (Michie,

2003). The debate between the Keynesian, monetary and post-Keynesian economists can

16

only be resolved by evaluating empirical evidence to conclude whether or not there is a

relationship between inflation and unemployment or whether the Phillips curve is vertical

or not. The empirical evidence is presented below.

INTERNATIONAL EMPIRICAL EVIDENCE

Existence of a Trade-Off

Empirical evidence supports the existence of a trade-off between inflation and

unemployment. Research-conducted for European countries (Holden, 2002); certain

member countries of the Organisation for Economic Corporation and Development (Ball,

Mankiw & Nordhaus, 1999) as well as for the United States of America (Karanassou,

Sala & Snower, 2010; Banarjee et al., 2008; Akerlof et al., 2000)- demonstrates that there

is a trade-off between inflation and unemployment. The trade-off results from, inter alia,

expectations about future inflation coupled with rigidities in nominal wages (Akerlof et

al., 1996; Akerlof et al., 2000); regulated and highly unionised labour markets (Holden,

2002) and shocks to economies (Svenssons, 1999). In their study, Akerlof et al. (2000)

find that employees’ expectations deviate from rational behaviour at low inflation levels

and that workers and employers tend to ignore the impact of low inflation when they

negotiate wages and prices. Akerlof et al. (2000:4) use survey data based on inflation

expectations of workers and employers to determine the inflation rate and the

unemployment level, which optimises the trade-off between inflation and unemployment.

In contrast to models based on rational expectations; Akerlof et al. (2000) find that the

Phillips curve is not vertical at low inflation levels, and that there is a trade-off between

inflation and unemployment.

In his study on European countries, Holden (2002) find that countries with

regulated labour markets, and where the majority of workers are unionised, face a trade-

off between inflation and unemployment when they target low inflation rates. This

situation arises because, in unionised labour markets, wages can only be changed through

mutual consent between employers and workers (Holden, 2002). Hence, workers and

their labour unions possess bargaining power, which they can use to prevent employers

from effecting cuts in nominal wages when inflation is low (Holden, 2002) The inability

of employers to reduce nominal wages unilaterally means that, in the long run, increases

in nominal wages lead to increases in inflation rate (Holden, 2002).

Determination of an Optimal Inflation Rate

The existence of a trade-off means that there is an optimal inflation target. An appropriate

inflation target would ideally be close to the optimal inflation rate on a nonlinear Phillips

curve. Given the presence of a trade-off, studies have been conducted in certain

developed countries (Akerlof et al., 2000; Fortin, 2001; Lundborg & Sacklen, 2006;

Maugeri, 2010; Wyplosz, 2000) to evaluate whether the inflation target is set at a level

that optimises the trade-off between inflation and unemployment. For example, Fortin

(2001:7) in his analysis of inflation in Canada over the period 1992 to 2002 find that the

inflation target, which is set by the Bank of Canada exhibits a trade-off between inflation

17

and unemployment. While the Bank of Canada has set an inflation rate band of 1%-3%,

Fortin (2001) find that an optimal inflation target band for Canada would be 2%-3%.

In a study of inflation targeting policy in Sweden, Lundborg & Sacklen (2006)

find that the inflation target of 2 percent, which is pursued by the European Central Bank,

is not at a level that optimises the trade-off between inflation and unemployment.

Lundborg & Sacklen (2006) analyse data from 1963 to 2000 and find that the trade -off

would be optimised if the European Central Bank targeted an inflation rate of 4 percent.

A similar study was conducted by Maugeri (2010) for Italy for the period 1960to2003.

Maugeri (2010) finds that the inflation target of 2 percent, which is pursued by the

European Central Bank, is not optimal for Italy. Maugeri notes that the inflation rate that

would minimise the trade-off between inflation and unemployment in Italy is between

15% and 20% (Maugeri, 2010). Making use of survey data on inflation expectations of

economic agents in regulated and unionised labour markets of France, Germany, the

Netherlands and Switzerland from 1960 to 1999, Wyplosz (2000) investigates the

inflation rate that optimises the trade-off of the two variables in these countries. Wyplosz

finds that the 2 percent inflation target, which is pursued by the European Central Bank,

results in trade-off between inflation and unemployment, that is, the low inflation target

leads to the increase in unemployment in these countries. The results from Wyplosz’s

study (2000) suggest that an optimal inflation rate for France, Germany, the Netherlands

and Switzerland should be 5 percent. While an optimal inflation target range should

refrain from generating excess unemployment, Mishkin and Westelius (2008) suggest

that the higher the uncertainty about the inflation process, the wider should the target

range be. This is true for emerging market countries, which are more vulnerable to

external shocks and likely to have more uncertainty about inflation outcomes.

Empirical Evidence in South Africa

Most of the literature on inflation targeting in South Africa (e.g. du Plessis and Burger,

2006; Fedderke and Schaling, 2005; Nell, 2002 and Pretorius and Smal, 1994) indicates

the existence of trade-off, in the short run, between inflation and unemployment or

between inflation and proxies for demand effects such as marginal costs and output gaps.

Early studies, which were conducted in the 1960s and 1970s, find trade-off between

inflation and unemployment (e.g. Gallaway, Koshal and Chapin (1970); Hume (1971);

Hodge (2002); du Plessis and Burger (2006)) - and between inflation and output gaps

(e.g. Krogh (1967); Truu (1975); Strydom and Steenkamp (1967);du Plessis and Burger

(2006)). Recent studies based on output gaps (Pretorius & Smal, 1994; Fedderke &

Schaling, 2005; Nell; 2002) mainly find evidence of trade-off between inflation and

output gaps. Studies, based on expectations-augmented Phillips curve of South Africa

(Pretorius and Smal, 1994 and Fedderke and Schaling, 2005), find that the trade-off

occurs indirectly through labour costs rather than through prices.

More recent studies (Leshoro, 2012; Phiri, 2010; Gupta and Uwilingiye, 2008)

demonstrate that the 3%-6% inflation target range limits the level of economic growth in

South Africa. Leshoro (2012) asserts that an inflation rate greater than 4 percent has a

negative effect on GDP growth rates. Gupta and Uwilingiye (2008) assert that a 3% to

6% inflation target range results in a welfare loss, which ranges from 0.34 percent to 0.67

percent of GDP. Phiri (2010:354) find an inflation threshold of 8 percent and concludes

18

that any inflation rate below and above the threshold will have an adverse effect on

growth.

Methodology

To determine the optimal level of inflation target in South Africa, this paper follows the

methodology developed by Ball and Mankiw (2002), which rests on the premise that

there is a short run trade-off between inflation and unemployment. Given the existence of

the short run trade-off, Ball and Mankiw (2002) show that there exists a level of

unemployment that is consistent with stable inflation. The unemployment level that

corresponds with a stable inflation is known as the non-accelerating inflation rate of

unemployment (NAIRU) (Gordon, 1997 and Ball & Mankiw, 2002). Thus, this paper

uses an expectations-augmented Phillips curve to estimate a time-varying NAIRU for

South Africa from 1980 to 2015. The NAIRU is not directly observable but is estimated

based on variables which are used to determine an expectations-augmented Phillips curve

(see Staiger, Stock, and Watson, 1997; Boone, Giorno, Meacci, Rae, Richardson and

Turner, 2003). To determine if the SARB target is set at a level that optimises the trade-

off between inflation and unemployment, the estimated NAIRU is used to determine a

stable inflation rate for South Africa. The inflation rate, based on the estimated NAIRU,

is then compared to the inflation target range that is adopted by the South African

Reserve Bank. .

The short-run trade-off between inflation and unemployment can be expressed

as:

U  (1)

Where 𝜋 represents the inflation rate; 𝑈 represents the unemployment rate;  is a parameter which represents a constant term in the equation; and  >0 is a parameter

which measures how the inflation rate responds to changes in the unemployment rate.

Nonetheless, the amended version of the inflation-unemployment dynamics is

represented as:

    *UUe (2)

Where e

 is expected inflation, *

U is the NAIRU and  is the supply shock. Equation 2 shows that the actual inflation depends on the expected inflation and how U deviates

from *

U . Supply shocks, such as oil crisis and changes in the exchange rate, may affect also the level of inflation.

Ball and Mankiw (2002:118) acknowledge that economic agents base their

decisions on adaptive expectations 1 . Thus, the expectations-augmented Phillips curve,

which is based on adaptive expectations, is thus presented as follows:

19

    

UU *

1 (3)

Where 𝜋−1 represents the inflation rate observed during the previous period. The expectations-augmented Phillips curve, as represented by equation 3 above,

is used as a base to derive an estimate of a time-varying NAIRU that is used to determine

a stable inflation rate. Ball and Mankiw (2002) show that to estimate the NAIRU,

Equation 3 should be rewritten as:

  UU *

(4)

Suppose that the value of parameter  is estimated, Equation 4 becomes:







 

 * UU (5)

Given that  and U are observed from the data, the left-hand side of Equation 5 can be

computed to provide the estimate of 

 

* U . Ball and Mankiw (2002) suggest the use

of Hodrick-Prescot (HP) filter to obtain *

U , which represents the longer-term trend and



 , the shorter-term supply shock or cyclical movement. The inflation rate, which

corresponds to the natural rate of unemployment, is estimated by imputing values for

parameters derived above into Equation 4.

Data and Empirical Results

In this study, we use the headline inflation rate ( )2 and the official unemployment rate based on the narrow definition ( U ) to evaluate the appropriateness of the current inflation target range. Quarterly data from 1980 to 2015 is used and data for  and U

are sourced from Quantec. In order to estimate the NAIRU ( *

U ) we estimate the

coefficient  from Equation 1. However, given that both  and U have a unit root, we have to test if there is a cointegrating relationship between these variables by making use

of the Engel-Granger cointegration test. It is important to note that the choice of the

Engle-Granger cointegration over other cointegration techniques, such as the Johansen

technique, is justified by the fact the Engle-Granger cointegration technique is robust

even when series are fractionally integrated or depict long memory (Gonzalo and Lee,

2000). Gonzalo and Lee (2000) show that it is very difficult to distinguish between series

20

that are fractionally integrated and those that have unit root (I(1)). By applying the Engle-

Granger cointegration technique, we believe that the results obtained will be robust even

though the series are fractionally integrated or have long memory. Moreover, Silvapulle

and Podivinski (2000) indicate that modellers should not be concerned by the possibility

of small departure from the condition of non-normality when using cointegration

technique even in finite samples. However, the authors emphasise that ARCH and

GARCH effects may be more problematic and compromise the results obtained from a

cointegration technique. The results of the Engle-Granger cointegration test are reflected

on Table 1. The results show that the null hypothesis of no cointegration between  and U is rejected when inflation is an endogenous variable.

The study adopts the results of the Engle-Granger cointegration test and

estimated the cointegrating parameter in the relationship between unemployment and

inflation. The estimation, as per Equation 1, yields a value of 0.66 for parameter  . The

estimated equation is represented as:

U66.02.23  (6)

TABLE 1. RESULTS OF THE ENGLE-GRANGER COINTEGRATION TEST

Note: the null hypothesis of no cointegration is rejected when  is endogenous

Given the above suggestion by Silvapulle and Podivinski (2000), we need to make sure

that there is no ARCH effect in the results obtained. Thus, we perform the

heteroscedasticity test to examine the presence of the ARCH effect in the obtained

results. Table 2 presents the results of the ARCH test for heteroscedasticity in the

residual. This test aims to detect the presence of time-varying volatility or ARCH effect

in the residual. The results presented in Table 2 show that the null hypothesis of no

ARCH is not rejected with the F-statistics and the observed R-Squared. This indicates

that the residual is homoscedastic and there is no need to account for time-varying

volatility in the estimated model.

TABLE 2. HETEROSCEDASTICITY TEST: NULL HYPOTHESIS OF NO ARCH

F-Statistics 1.707

Probability 0.2001

Obs R-Squared 1.724

Probability Chi-Square 0.1892 Note: The probabilities show the rejection of the null hypothesis

Knowing the value of  , we proceed to estimate the natural rate of unemployment as per the procedure described above by using the HP filter. Figure 1 below provides a

diagrammatic representation of South Africa’s NAIRU, together with the headline

inflation and actual unemployment rate, from 1980 to 2015.

Variable z-statistic Probability  -29.6342 0.0214

U -18.20813 0.2031

21

FIGURE 1. GRAPHICAL REPRESENTATION OF SOUTH AFRICA'S

ESTIMATED NAIRU, INFLATION AND UNEMPLOYMENT FROM 1980 TO

2015

Figure 1 shows that the estimated time-varying NAIRU increases substantially from

single digit figures in the 1980s to double digit figures from the late 1980s to 2000. From

the year 2000 to 2015, the NAIRU stabilises around 23 to 25 percent. In Kabundi et al.

(2015) and Viegi (2015) we find support for these findings as both studies show that the

NAIRU stabilises around these rates in South Africa. Figure 1 further shows that the

unemployment rate is below the NAIRU during the period 1994 to 1997. This period

corresponds to increasing economic activities in South Africa. However, the

unemployment rate in South Africa trended above the NAIRU during the period 1998-

2000 marking the period of contagion from the Russian and Latin American financial

crises and during the period post 2008, signifying the effects of global financial crisis o n

the South African labour market. The trend of unemployment compared to the NAIRU is

an evidence of the contribution of external shocks to the labour market and economic

activities in South Africa.

It is clear from Figure 1 that the relationship between inflation ( ), unemployment and NAIRU is far from supporting the evidence of a vertical Phillips

curve in South Africa, especially after 2004. Figure 1 shows that deviation of

unemployment rate from NAIRU translates to changes in inflation rate. For example, in

the period 2004-2008 the unemployment rate in South Africa is below the NAIRU. This

deviation of unemployment rate from the NAIRU coincides with the increase in inflation

rate in South Africa. However, during the period of global financial crisis and the

afterward of the crisis, unemployment soars above the NAIRU as inflation rate decreases.

This finding indicates thatthe difference between the NAIRU and unemployment

coincide with the change in inflation in South Africa as per the theory of the augmented

Phillips curve. This occurrence indicates that actual inflation is not only influenced by

inflation expectations but also by the deviation of the unemployment rate from the

NAIRU. This further shows that the Phillips curve is not vertical in South Africa, as

actual inflation does not depend solely on inflation expectations. The final step of the

analysis is to obtain estimates of the stable inflation rate that would minimise the trade-

8

12

16

20

24

28 0

4

8

12

16

20

1980 1985 1990 1995 2000 2005 2010 2015

inflation

unemployment

NAIRU

22

off between inflation and unemployment. The estimates from Equation 3 yield a rate of

inflation, which ranges from 1.4 percent to 11.5 percent from 2000 to 2015. Figure 2

compares the actual inflation rate (  ) and the inflation rate derived from the South African NAIRU ( InflaN ).

FIGURE 2. SOUTH AFRICA'S INFLATION RANGE BASED ON ESTIMATES

OF NATURAL RATE OF UNEMPLOYMENT (INFLAN) AND THE ACTUAL

INFLATION RATE

0

4

8

12

16

20

1980 1985 1990 1995 2000 2005 2010 2015

Actual inflation InflaN

The results show that InflaN and Actual inflation (  ) have similar trends, but, in

spite of this, they have different ranges. Moreover, these results indicate that, if South

Africa were to put in place an inflation target range based on the NAIRU, it would have

to target an inflation rate that ranges from 1.4 to 11.5 percent. This range is different to

the official inflation target of 3% to 6% adopted by the South African Reserve Bank

(SARB). It is a reality that when South Africa implemented inflation targeting, it chose

an inflation range that coincided with its trading partners rather than a range that could

optimise the trade-off between unemployment and inflation.

It is clear that monetary policy makers in South Africa are adopting the inflation

targeting policy on the premise of a vertical Phillips Curve without sound evaluation of

the domestic context and reality. It is not surprising that stakeholders such as the trade

unions continue to believe that the South African Reserve Bank has been too restrictive in

the conduct of monetary policy. Although South Africa’s trading partners have set

inflation targets at similarly lower rates, these countries have lower unemployment rate

and NAIRU than South Africa’s which gives them the leverage to target inflation rates

over narrow bands. Table 3 indicates that a vast majority of these countries have low and

stable inflation rate interval, and with the exception of Brazil, all the countries, Thailand,

Peru and Israel target inflation at rates which are slightly below South Africa’s inflation

target range. It is then difficult to understand why South Africa has a lower upper bound

inflation rate and a tighter inflation target interval than one of his important trade partner,

Brazil. A wider inflation target, like the one of Brazil is ideal for emerging market

economies that are vulnerable to external shocks. A wider inflation target range could

prevent monetary authorities from frequently reacting to external shocks. Stringent anti-

inflationary policies might have caused persistent and high unemployment in South

Africa. Literature shows that a number of emerging market economies have harmed their

economies from unnecessarily reacting to external shocks. For example, Mackowiak

(2007) shows that United States (US) monetary policy shocks affect a larger fraction of

23

the variance in the aggregate price level and aggregate output in emerging economies

than of the variances in the same variables in the US itself. Moreover, Kaminsky et al.

(2005) indicate that emerging market economies are overwhelmingly procyclical in their

conduct of monetary policy and often deepen downturns when reacting to external

shocks. South African monetary authority may have caused reduced economic activities

as it responds to the effects of external shocks on the domestic economy with restrictive

measures. Indeed a number of studies have found that restrictive monetary policy

reactions often fail to reduce inflation in South Africa (Bonga-Bonga and Kabundi,

2011). Moreover, in a historical perspective, the 3% to 6% inflation range adopted by the

SARB is strikingly low, as Figure 1 shows that in the 1980s when South Africa had one

digit unemployment rate, inflation rate was between 12% and 14%.

TABLE 3. EMERGING MARKETS' SOUTH AFRICA'S COMPETITORS WHO

ADOPTED INFLATION TARGETING

Country Inflation

Target range

Unemployment

rate

Chile 2 - 4% 5.90%

Israel 1 -3 % 6.20%

Brazil 1.5 - 8.5% 6.50%

Czech Republic 2 - 4% 7%

Thailand 0.90% 0.80%

Source: Levin, Natal & Piger (2004) and Fraga, Goldfajn & Minella (2003), World Bank (2015)

Note: unemployment rate is for the period 2006-2010.

CONCLUSIONS

This paper aims to assess whether the SARB 3% to 6% inflation target range is at a rate

that optimises the trade-off between inflation and unemployment. Empirical work shows

that countries with low levels of inflation and/or whose Phillips Curve is vertical (many

of which are developed economies) may adopt low inflation rate/ ranges targets. Studies

also show that despite South Africa having non-vertical Phillips Curve, significant high

levels of unemployment, and high-income inequality, its inflation targeting policy is

modelled around a narrow inflation rate band with a lower upper bound, unlike countries

at similar level of development such as Brazil. Furthermore, while South Africa’s

inflation target range compares well with that of other emerging market countries, the

level of unemployment in South Africa is higher than that of these emerging market

countries. In this research work, we estimate the stable inflation rate based on the

expectations-augmented Phillips curve for South Africa. Given the magnitude of the

computed NAIRU, the estimation results provide an inflation range that is wider than the

current inflation target range set by the South African Reserve Bank. The paper

concludes that the current tighter inflation target policy in South Africa is based on

unsound fundamentals and may have led to some unnecessary responses by the monetary

authority to apply restrictive measures which in turn have been detrimental to output

growth and reduced employment, among other things.

The policy implication of this paper is that the South African Reserve Bank

(SARB) should rethink the current inflation target range of 3% to 6%. Such a narrow

24

range of inflation target could have a negative effect on output and unemployment in the

country given that monetary authority may be forced to apply unnecessary contractionary

monetary policy. Moreover, the findings of this paper that SARB should broaden the

inflation target range could even suggest that dual targets, inflation and output targets,

may be appropriate for South Africa. The poor output growth that South Africa

experiences warrants scrutiny by policymakers.

ENDNOTES

1 Ehlers and Steinbach (2007) show that economic agents in South Africa make use of adaptive

expectations to a certain extent in forming their expectations of future inflation. 2 It is important to note that the South African Reserve Bank (SARB) was targeting CPIX-inflation

from 2000 t0 2008 instead of the headline CPI. In 2008, the SARB reverted to targeting headline

CPI inflation.

REFERENCES

Arkerlof, GA, Dickens, WT, Perry, GL, Gordon, RJ, Mankiw, NG (1996) The

macroeconomics of low inflation. Brookings Papers on Economic Activity 1999: 1 – 76.

Arkerlof, GA, Dickens, WT, Perry, GL, Bewley, TF, Blinder, AS (2000) Near-

Rational Wage and Price Setting and the Long-Run Phillips Curve. Brookings Papers on

Economic Activity 2000: 1 – 60.

Ball, L, Mankiw, NG, Nordhaus, WD (1999) Aggregate Demand and Long-run

Unemployment. Brookings Papers on Economic Activity 1999: 189 – 251.

Ball, L, Mankiw, NG (2002) The NAIRU in theory and practice. Journal of

Economic Perspectives 16: 115 – 136.

Banerjee, A, Galioni, S, Levinsohn, J, McLaren, Z, Woolard, I (2007) Why has

unemployment risen in the new South Africa? National Bureau of Economic Research,

NBER Working Paper Series, Working Paper No. 13167.

Banerjee, A, Galioni, S, Levinsohn, J, McLaren, Z, Woolard, I. (2008). Why has

unemployment risen in the new South Africa? Economics of Transition 16: 715 - 740.

Bernanke, BS, Laubach, TS, Mishkin, FS (1999) Inflation targeting. Princeton

University Press, Princeton

Bernanke, BS (2003) Panel discussion, October 17. www.federalreserve.

gov/boarddocs/speeches/2003. Accessed 12 Decembre 2015.

Bonga-Bonga, L, Kabundi, A (2011) Monetary policy action and inflation in

Siuth aFrica: an empirical analysis, African Finance Analysis 13: 25-37.

Boone, L, Giorno, C, Meacci, M., Rae, D, Richardson, P, Turner, D (2003)

Estimating the structural rate of unemployment for the OECD countries. OECD

Economic Studies 2002/2: 172 - 211.

Burger, P, Marinkov, M (2006) The South African Phillips curve: How

applicable is the Gordon model? South African Journal of Economics 74: 172 – 189.

Casteleijn, AJH (1999) The Viability of Implementing an Inflation Targeting

Monetary Policy Framework in South Africa. South African Reserve Bank: Pretoria.

Department of Economic Development (2010). New Growth Path Framework.

Department of Economic Development: Pretoria Department of Finance (1996) Growth,

25

Employment and Redistribution: A Macroeconomic Strategy. Department of Finance:

Pretoria.

Du Plessis, S, Burger, R (2006) A New Keynesian Phillips Curve for South

Africa. Paper presented at the South African Reserve Bank Conference: Pretoria.

Ehler, N, Steinbach, R (2007) The formation of inflation expectations in South

Africa. Working paper no. 3243, South African Reserve Bank.

Erceg, CJ (2002) Choice of an Inflation Target Range in a Small Open

Economy. AEA Papers and Proceedings 92: 85 – 89.

Fedderke, JW, Schaling, E ( 2005) Modelling Inflation in South Africa: A

multivariate cointegration analysis. South African Journal of Economics 73: 79 - 92.

Fischer, S (1996) Why are central banks pursuing long-run price stability?

Federal Reserve Bank of Kansas City: Kansas City.

Fortin, P, (2001) Inflation targeting: the three percent solution. Policy Matters 2:

1 – 16.

Fraga, A, Goldfajn, I, Minella, A (2003) Inflation targeting in emerging market

economies. National Bureau of Economic Research, NBER Working Paper Series,

Working Paper No. 10019.

Friedman, M, (1968) The Role of Monetary Policy. The American Economic

Review 58:1 - 17.

Gonzalo, J and Lee, TH (2000). On the robustness of cointegration tests when

series are fractionally integrated. Journal of Applied Statistics, 27(7): 821-827.

Gordon, RJ (1997) The time-varying NAIRU and its implications for economic

policy. Journal of Economic Perspectives 11:11 – 32.

Gupta, R, Uwilingiye, J (2008). Measuring the Welfare Cost of Inflation in

South Africa. Economic Research Southern Africa, Working Paper No. 68.

Hodge, D (2002) Inflation versus unemployment in South Africa: Is there a

trade-off? South African Journal of Economics 70: 418 – 443.

Holden, S (2002) The costs of price stability – downward nominal wage rigidity

in Europe. The National Bureau of Economic Research, Working Paper No. 8865.

Hsing, Y (2009) Estimating the time-varying NAIRU for Germany and policy

implications. Applied Economic Letters 16:469 – 473.

Kabundi, A, Schaling, E, Some, M (2015) Estimating a Phillips Curve for South

Africa: A Bounded Random Walk Approach. ERSA working paper no. 568.

Kahn, B (2008) Challenges of Inflation Targeting for Emerging-market

Economies: The South African Case. Paper presented for the South African Reserve Bank

Conference Series on Challenges for monetary policy-makers in emerging markets,

Pretoria.

Kaminsky, GL, Reinhart, CM, Vegh, CA (2005) When it rains, it pours:

procyclical capital flow and macroeconomic policies, in NBER macroeconomic Annual

2004, Vol.19, ed. By Gertler, M. and Rogoff, K., Cambridge: NBER.

Karanassou, M, Sala, H, Snower, DJ (2010) Phillips curves and unemployment

dynamics: A critique and holistic perspective. Journal of Economic Perspectives 24: 1 -

51.

Klasen, S, Woolard, I. (2008) Surviving Unemployment without State Support:

Unemployment and Household Formation in South Africa. Journal of African Economies

18: 1 – 51.

26

Leshoro, TLA. (2012) Estimating the inflation threshold for South Africa.

Economic Research Southern Africa. ERSA, Working Paper 285.

Levin, AT, Natalucci, FM, Piger, JM (2004) The macroeconomics effects of

inflation targeting. Conference proceeding of the 28 th

annual economic policy conference

on “Inflation Targeting: Prospects and Problems” held in St Loius. Conducted by the

Federal Reserve Bank of St Louis. St Louis: Federal Reserve Bank of St Louis.

Lundborg, P, Sacklẻn, H (2006) Low-inflation targeting and long run

unemployment. The Scandinavian Journal of Economics 108:397 – 418.

Mackowiak, B (2007) External shocks, US monetary policy shocks and

macroeconomic fluctuations in emerging market. Journal of monetary economics 54:

2512-2520.

Mankiw, NG (2001) The Inexorable and Mysterious Trade-off between Inflation

and Unemployment. The Economic Journal 111: C45 – C61.

Maugeri, N (2010) Macroeconomic Implications of Near Rational Behaviour:

An Application to the Italian Phillips Curve. Unpublished doctoral thesis. Siena:

Università di Siena.

Mboweni, TT (1999) Inflation targeting in South Africa. The South African

Journal of Economics 67: 401 - 409.

McCord, A, Bhorat, H (2003). Employment and Labour Market Trends. In:

Kraak A, Perold, H (eds) Human Resources Development Review. HSRC Press: Cape

Town

Michie, J (2003) The Institutional Underpinnings of the Unemployment-Inflation

Relationship: A Review Paper. Human Sciences Research Council: Pretoria:

Mishkin, FS, Westelius, NJ (2008) Inflation Band Targeting and Optimal

Inflation Contracts. Journal of Money, Credit and Banking 40: 557-582

National Labour and Economic Development Institute (2004). Highlights of the

current labour market conditions in South Africa. Johannesburg: National Labour and

Economic Development Institute.

Nell, KS (2002) Is low inflation a precondition for faster growth? The case of

South Africa. University of Kent, Department of Economics, Working Paper No. 0011.

Padayachee, V (2001) Central bank transformation in a globalized world: The

Reserve Bank in post-apartheid South Africa. Journal of International Development 13:

741 – 765.

Phillips, AW (1958) The Relation between Unemployment and the Rate of

Change of Money Wage Rates in the United Kingdom, 1861 – 1957. Economica 25:283-

299.

Phiri, A (2010) At what level is inflation least detrimental towards finance-

growth activity in South Africa? Journal of Sustainable Development in Africa 12: 354 –

364.

Pretorius, CJ and Smal, MM (1994) A macro-economic examination of the

price-formation process in the South African economy. The South African Reserve Bank

Quarterly Bulletin 191: 25 – 36.

Ricci, LA (2005) Bringing inflation under control. In: Nowak, M, Ricci, LA

(eds) Post-Apartheid South Africa: The First Ten Years. International Monetary Fund:

Washington.

27

Rossouw, J, Padayachee, V (2008) An Analysis of inflation from a central

banking perspective: The South African experience since 1921. University of Kwazulu

Natal: School of Development Studies. Working Paper No. 50.

Silvapulle, PS and Podivinsky, JM (2000). The effect of non-normal

disturbances and conditional heteroskedasticity on multiple cointegration tests. Journal of

Statistical Computation and Simulation, 65(1-4): 173-189

Staiger, D, Stock, JH, Watson, MW (1997) How Precise are Estimates of the

Natural Rate of Unemployment? The Journal of Economic Perspectives 11: 34 – 49.

Svenssons, LEO (1999) Price-level targeting versus inflation targeting: A free

lunch? Journal of Money, Credit and Banking 31:277 – 295.

Van der Merve, EJ (2004) Inflation targeting in South Africa. The South African

Reserve Bank Occasional Paper No. 19.

Viegi, N (2015) Labour Market and Monetary Policy in South Africa. SARB

Working Paper 15/01.

World Bank (2015) World Development Indicators. http://beta.worldbank.org.

Accessed 27 january 2016.

Wyplosz, C (2000) Do we know how low should inflation be? Conference

proceeding of the first Central Banking Conference on “Why Price Stability” held in

Geneva. Conducted by the European Central Bank. Geneva: Graduate Institute of

International Studies.

Copyright of Journal of Developing Areas is the property of Tennessee State University, College of Business and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.