Testing for a common Phillips curve in common monetary area of Southern Africa

Abstract

PurposeThis paper sets out to investigate whether the four members of the common monetary area (CMA) regime experience similar inflation-unemployment dynamics as explained by the Phillips Curve phenomenon.Design/methodology/approachThis study uses a combination of seemingly unrelated regression (SUR) and Copula based marginal regression techniques to investigate existence of a common Phillips curve (PC) between members of the CMA. Model estimation was done using country specific annual time series data for inflation, unemployment and imports spanning from 1980 to 2014.FindingsWe find evidence of contemporaneous correlation between the residuals of individual CMA PC equations and a statistically significant trade-off between inflation and unemployment for all CMA countries. Wald test results of cross-equation restrictions reveal a 9.94% chance of a common unemployment coefficient for CMA countries.Originality/valueTogether, the results of the SUR and Gaussian Copula techniques provide mixed and inconclusive evidence to support the existence of a common PC among CMA member states. This study is the first of its kind in examining this phenomenon for currency board regimes like CMA, and one of the very few among emerging market economies.

Evaluation of inflation and unemployment and plotting the Philips curve of the Czech Republic

“…Subsequently, I found that Álvarez and Correa-Lopéz [15] used a model with gradual regressions for their prognosis and evaluated their results using meta-regression. In addition, Damane and Sifat [16] used a combination of seemingly unrelated regression (SUR) and marginal regression techniques based on a multidimensional cumulative function to investigate the existence of a common Phillips curve between CMA members. Similarly, Furtula, Durkalic and Simionescu [17] examined the paradigm of the Phillips curve in macroeconomics and examined the relationship between inflation and the unemployment rate, which they analyzed on empirical data using Bayesian linear regression models.…”

“…His algorithm creates a regression model of expected inflation using a large panel of macroeconomic data as possible variables. However, for example, Daman's and Sifat's [16] model estimates were made using data on the annual time series of specific countries for inflation, unemployment and imports. Tule, Salisu and Chiemeke [19] also propose a Phillips curve based on supply for an oil-dependent (Nigerian) economy.…”