Re‐Examining the Role of Structural Change and Nonlinearities in a Phillips Curve Model for South Africa

• DOI: http://doi.org/10.1111/saje.12187
• Date: 01 June 2018
• Author: Kevin S. Nell
• Published date: 01 June 2018

RE-EXAMINING THE ROLE OF STRUCTURAL CHANGE

AND NONLINEARITIES IN A PHILLIPS CURVE MODEL FOR

SOUTH AFRICA

KEVIN S.NELL*

Abstract

Although studies generally find evidence of a Phillips curve-type relationship in South Africa,

uncertainty remains about the relevance of the model over a relatively long sample period, and

whether conventional output gap measures are suitable proxies for demand pressure. This paper

reviews research which shows that the Phillips curve model prevails over an extended sample,

provided that the benchmark specifications include major structural changes in the balance-of-

payments and labour market, and account for shifts in the root causes of inflation. When this is

done, a linear specification with an output gap in levels correctly predicts the non-trended infla-

tion pattern over the period 1971(Q1)–1984(Q4), whereas a piecewise concave curve with an

output gap in growth rates accurately forecasts the decelerating inflation pattern during

1986(Q1)–2001(Q2). A novel feature of the concave model is that it remains statistically robust

and structurally stable when it is estimated until 2015(Q4). The concave model imparts a disin-

flationary bias, which suggests that monetary policy should be more expansionary during down-

swing phases of the business cycle and neutral during upswing phases. The analysis also considers

how the shape of the Phillips curve might change if the balance-of-payments constraint on

demand is relaxed in a significant way.

JEL Classification: C22, E31, E52, O11, O23

Keywords: Phillips curve, labour market, balance-of-payments, nonlinear, structural inflation, cost-

push inflation, demand-pull inflation, inflation targeting

1. INTRODUCTION

A growing number of studies find that the conventional inflation-output gap Phillips

curve relationship in South Africa no longer holds over an extended sample period, espe-

cially since the 1970s (see, e.g.Akinboadeet al., 2002; Fedderke and Schaling, 2005;

Burger and Marinkov, 2006; Hodge, 2006; Burger and Du Plessis, 2013; Fedderke and

Liu, 2016; Leshoro and Kollamparambil, 2016; Phiri, 2016). The output gap in these

studies is either insignificant or contains the incorrect theoretical sign.

1

In an attempt to

recover the missing Phillips curve, some researchers have focused on alternative proxies

* Corresponding author: Associate Professor, College of Business and Economics, School of

Economics, University of Johannesburg, Corner Kingsway and University Road, Auckland

Park 2006, Johannesburg, South Africa. E-mail: knell@uj.ac.za

1

Although Phiri (2016) claims that the conventional inflation-output gap relation holds in one

of his Phillips curve specifications over the period 1970–2014, the output gap is marginally signifi-

cant and carries a negative sign (see table 5, equation 3.2 in Phiri). This is inconsistent with the the-

oretical prediction of the conventional Phillips curve model, which predicts a positive relation

between the inflation rate and output gap.

V

C2018 Economic Society of South Africa. doi: 10.1111/saje.12187

173

South African Journal of Economics Vol. 86:2 June 2018

South African Journal

of Economics

for the output gap. The basic premise, as argued in Burger and Du Plessis (2013) and

Fedderke and Liu (2016), is that potential output in the constructed output gap variable

is poorly proxied by the conventional Hodrick–Prescott (HP) filter (Hodrick and

Prescott, 1980).

2

The alternative measures of demand pressure, such as the derived mar-

ginal cost variable in Burger and Du Plessis’s (2013) New Keynesian Phillips Curve

model and the growth rate of the M2 money supply and government expenditure in

Fedderke and Liu (2016), perform relatively well, with statistically significant and

theory-consistent signs in the inflation rate equations.

On the other hand, Reid and Du Rand’s (2015) micro-founded Sticky Information

Phillips Curve (SIPC) model, with the output gap as a proxy for demand pressure, fits

the data well over the period 2000(Q3)–2010(Q4). In fact, they show that the output

gap derived from the HP filter appears to be a better proxy for demand pressure than

their constructed marginal cost variable.

The relevance of the output gap, as opposed to other proxies for demand pressure, is an

important issue for the conduct of modern-day monetary policy. One of the main transmis-

sion mechanisms that underlies the South African Reserve Bank’s (SARB’s) monetary policy

framework and, indeed, most central banks worldwide, is a direct link from the monetary

policy instrument (repo rate) to output (demand) and the inflation rate (Svensson, 1999,

2000; Smal and De Jager, 2001; Rudebusch and Svensson, 2002; Nell, 2004). In this con-

text, the Factor Vector Autoregressive methodology employed by Kabundi and Ngwenya

(2011), using monthly data from February 1985 to November 2007, suggests that there

may be a link from the repo rate to the capacity utilisation ratio (or the output gap if the

ratioisexpressedinlogs)andtheconsumerpriceinflationrateinSouthAfrica.

Although studies over a more recent sample period support the inflation-output gap

nexus of the Phillips curve model in South Africa, uncertainty still remains about the rele-

vance of the relationship over a longer sample period, and whether the output gap, derived

from conventional measures, is a suitable proxy for demand pressure. The difficulties in

identifying a robust Phillips curve model over time are not unique to a developing country

like South Africa, but have also been experienced in more advanced economies (Matheson

et al., 2013). The main purpose of this paper is to re-examine the inflation-output gap

nexus in South Africa over a relatively long sample period from 1971 to 2015. Following

the recent trend in the literature (Matheson et al., 2013; Coibion and Gorodnichenko,

2015), a stern test of the Phillips curve model is to examine how well the pressure of

demand can explain different inflationary episodes relative to other structural factors.

Accordingly, the first part of this paper provides a review of the main findings in Nell

(2006). The benchmark model in this study shows that the output gap derived from the

HP filter is statistically insignificant over the full sample period from 1971(Q1) to

2001(Q2). The model is then augmented for two major structural changes that occurred

in the South African economy. To capture labour market rigidities that became more

prominent since the mid-1980, the specifications consider the growth rate of the output

gap instead of the level across different regimes. Consistent with insider–outsider models,

the growth rate of the output gap measures downward rigidity in wages. In this frame-

work, the level of output/unemployment must continually fall/rise before wage demands

2

A smoothing technique of potential output, such as the HP filter, may not accurately capture

time-varying productivity shocks (Burger and Du Plessis, 2013).

174 South African Journal of Economics Vol. 86:2 June 2018

V

C2018 Economic Society of South Africa.