Causality relationship between energy consumption, economic growth, FDI, and globalization in SSA countries: a symbolic
- PDF / 239,528 Bytes
- 6 Pages / 595.276 x 790.866 pts Page_size
- 72 Downloads / 201 Views
SHORT RESEARCH AND DISCUSSION ARTICLE
Causality relationship between energy consumption, economic growth, FDI, and globalization in SSA countries: a symbolic transfer entropy analysis Seyi Saint Akadiri 1
&
Ahdi Noomen Ajmi 2,3
Received: 28 August 2020 / Accepted: 5 October 2020 # Springer-Verlag GmbH Germany, part of Springer Nature 2020
Abstract To substantiate the validity and usefulness of the symbolic transfer entropy test for longitudinal data, and how it validate or contrast existing study results generated using other forms of causality tests, we empirically examine panel-based causality relationships among foreign direct investment, energy consumption, globalization, and economic growth respectively, between the periods 1970 and 2014 using sub-Saharan African countries as a case study. Based on our findings, we are of the opinion that STE causality test results resonate existing findings, and it is a suitable causality approach for longitudinal data and for developing countries with poor-quality data, most specifically for the sample region. Keywords Symbolic transfer entropy . Panel data . Causality . Sub-Saharan Africa
Introduction Different causality tests proposed in the context of panel time series data have a common point that a linear autoregressive representation of the time series is required (Camacho et al. 2020). Based on Monte Carlo simulations, Camacho et al. (2020) provide different scenarios for which the size and power of these tests are seriously deteriorated: (i) if the linearity assumption is not verified, (ii) if there are structural breaks or extreme observations in some of the cross section units, (iii) if the datagenerating process is heterogeneous across the cross section units, (iv) if the causal dependence appears in the conditional variance, and (v) if we have qualitative data. To overcome this drawback, Camacho et al. (2020) introduced a non-parametric Responsible Editor: Nicholas Apergis * Seyi Saint Akadiri [email protected] Ahdi Noomen Ajmi [email protected] 1
Research Department, Central Bank of Nigeria, Abuja, Nigeria
2
College of Science and Humanities in Slayel, Prince Sattam bin Abdulaziz University, Al-Kharj, Kingdom of Saudi Arabia
3
ESC de Tunis, Manouba University, Manouba, Tunisia
Granger causality test procedure for unbalanced panel dataset on the concept of transfer entropy and a multiple-unit symbolic dynamics. In other words, the Camacho et al. (2020) test avoids the need to rely on a linear parametric representation of the dataset by translating the information set of the time series dynamics into symbols, using a simple symbolization technique based on Matilla-García et al. (2014). In addition, Camacho et al. (2020) causality test specified the null hypothesis of noncausality across the cross-unit sections by a simple stack of the symbols across the different cross units. Consequently, the test displays correct size and higher power in those cases where causality tests based in linear panel data specifications fail. The symbolic transfe
Data Loading...
