Are fiscal budgets sustainable in South Africa? Evidence from provincial level data

Abstract

This study uses the nonlinear autoregressive distributive lag (N-ARDL) model to investigate the expenditure-revenue relationship for all nine South African provinces using annual data spanning from 2000 to 2016. Whereas other cointegration models can only depict whether budgets are sustainable or not, the N-ARDL model presents features which further enable us to predict a course of action which individual provincial governments can take towards attaining higher levels of budgetary sustainability in both the short and the long-run. Ultimately, our empirical study demonstrates that the 'one rule fit all' strategy as suggested by previous studies may not be an appropriate approach seeing that provincial governments have differing requirements for attaining improved levels of budget sustainability JEL Classifications: C13, C32, H61, H72

Investigating the Government Revenue–expenditure Nexus: Empirical Evidence for the Free State Province in a Multivariate Model

“…For robustness, we also perform a multiple breakpoint unit root tests based on sequential breaks, recursive partitions and global crisis induced breaks proposed by Bai and Perron (2003) 23 . In general, structural breaks determined by all the unit-root test, are quite similar 24 , suggesting break dates for the period 2009 Q1 and 2011 Q2 (for government revenue variable); 2006 Q1 and 2006 Q2 (for government expenditure variable); 2009 Q2 and 2017 Q2 (for the real GDP variable); and 2010 Q3, 2016 Q1 and 2015 Q2 (for the inflation series). On the basis, the estimated dummy variables included in the error correction models consist of: (i) 2008 Q1-2012 Q2, capturing the full impact of the latest global economic recession and financial contagion; (ii) 2009 Q1-2011 Q4, which accounts for the period of a synchronized economic downturn in Africa; (iii) 2009 Q1-2010 Q4, accounting for the sharp fall in economic activity level in South Africa, during the latest global economic crisis period, prior a rebound at the beginning of 2011; and (iv) 2013 Q1-2018 Q4 to capture the implemented strict fiscal consolidation period to enforce prudent financial management and good governance across the three spheres of government (i.e., national, provincial, and municipal levels), since 2013.…”

“…On the methodological front, the econometric method and model specification employed in this study circumvent the shortcomings of earlier studies such as using bivariate modeling approach, low frequency (annual) data and lack of testing for structural breaks in used data (see, e.g., Ndahiriwe and Gupta 2010; Payne 2003; Islam 2001), which usually mask the intertemporal links between variables, and estimated model suffer from omitted variable problems. Most closely related to our work is Kavase and Phiri (2018), who investigate the fiscal sustainability across the nine South African provinces focusing on the existing asymmetric relationship between government revenue -expenditure. But, drawn inferences and conclusion of this particular study can be considered as unreliable since these authors used a bivariate model and high-frequency (annual) data that covers a short period (i.e., 2000 to 2016).…”

“…This test is carried out using Eviews 10, and the results are available upon request from the author. 24 The three unit root tests employed produce break dates, which mostly coincide with, for example (i) the latest 2007/8 global economic recession; (ii) prevailing weak economic condition in the country, since 2014; (iii) the adoption of strict fiscal consolidation process by the national government since 2014, to reduce wasteful public expenditure and the rapidly expanding fiscal debt (% of GDP) which is above 58%; and (iv) the gradual decline in generated revenue, since 2017 (see, Figure 3).…”

Do fiscal regimes matter for fiscal sustainability in South Africa? A Markov-switching approach

“…The results of these empirical studies on fiscal sustainability in South Africa, as listed in Table 1, vary depending on the study period and methodology applied. A plethora of studies in the literature used a linear approach Ganyaupfu 2014;Ganyaupfu and Robinson 2019;Ghartey 2010;Jibao et al 2012;Kavase and Phiri 2018;Lusinyan and Thornton 2007;Narayan and Nayaran 2006;Ndahiriwe and Gupta 2010;Nyamongo et al 2007;Siebrits et al 2014). Other studies (Baharumshah et al 2016;Burger and Marinkov 2012;Naraidoo and Raputsoane 2015;Phiri 2019) used non-linear techniques and found sustainable fiscal policy, while the empirical works of Cronje (1995), Roux (1993), and Schoeman (1994) investigated the fiscal sustainability of South Africa and found unsustainable fiscal policy.…”

“…Kavase and Phiri (2018) analysed fiscal budget sustainability in nine South African provinces over the period from 2000 to 2016 by utilizing a non-linear autoregressive distributive lag (N-ARDL) model to investigate the expenditure-revenue relationship. The study recommended budget reductions for the Eastern Cape, Northern Cape, Free-Sate, and KwaZulu-…”