Non-linear ADRL estimation of corruption and FDI inflow to Ghana
| DOI | https://doi.org/10.1108/JFC-05-2021-0106 |
| Published date | 11 August 2021 |
| Date | 11 August 2021 |
| Pages | 1042-1063 |
| Subject Matter | Accounting & finance,Financial risk/company failure,Financial crime |
| Author | Randolph Nsor-Ambala,Cephas Paa Kwasi Coffie |
Non-linear ADRL estimation of
corruption and FDI inflow to Ghana
Randolph Nsor-Ambala
Accounting and Finance Department, Ghana Institute of Management and Public
Administration, Accra, Ghana, and
Cephas Paa Kwasi Coffie
School of Management and Economics, University of Electronic Science and
Technology of China, Chengdu, China
Abstract
Purpose –This paper aims to examinethe effect of corruption on foreign direct investment (FDI) inflowin
Ghana. This provides answers to the call for further empirical examination of the contextual impact of
corruptionon FDI inflow.
Design/methodology/approach –The study uses a non-linearADRL time series econometric model to
estimatedata from the World Bank and the international country risk guide (1984–2019).
Findings –The study confirms the sand in the wheel and the grabbing hand hypothesis of the impact of
control of corruption(CoC) on FDI both in the short and long run. However, degradation on the CoC index has
a significant and more than a proportionateconstraint on FDI inflows, while an improvement in CoC has no
significant impacton improving FDI inflows. An explanation for this outcome was proposed aftercomparing
this finding to a similar prior study with a Nigerian data set (Zangina and Hassan, 2020). The proposed
explanationrelied mainly on the rational expectation hypothesis and drawingelements of the efficient market
hypothesis. FDI inflowsdo not react to outcomes or trajectories reasonably expected because such rationally
expected future outcomeswill have been modelled into existing FDI movement decisions.Instead, FDI flows
react to “surprises”andoften respond in a more than proportional manner.
Practical implications –Political leadership in Ghanashould be conscious of the severe adverse effects
of inaction or ineffective action in curbing corruption, leading to slippering in CoC rankings. In the case of
Ghana, the dependence of FDI on CoC is even more pronounced as the other variables within the specified
model show an insignificant impact on FDI. Additionally, admittedly aggregated cross-country data in
econometric modelling is appealingand has some empirical basis, but these must not erode the relevance of
country-specificstudies as both are needed to support theorization.
Originality/value –The paper is among the first to test for the asymmetric relationship between
corruption or its control thereof and FDI with a time series approach, and hence, the findings offer new
insight.
Keywords Corruption, Ghana, Foreign direct investments,
Non-linear autoregressive distributed lag, Rational expectation, Control of corruption
Paper type Research paper
1. Introduction and background
Despite the growing academic interest in the mechanisms, impact and outcomes of foreign
direct investments (FDI hereafter) flows within the development and macroeconomic
literature, contentionsremain regarding the impact of corruption on FDI inflow (Wang et al.,
2020). This paper examines the effect of corruption on FDI inflow, using a data set from
Ghana, answering the call for further empirical examination of the contextual impact of
corruption or its control thereonFDI inflow (Zangina and Hassan, 2020;Cie
slik and Goczek,
2018). The empirical evidence of the impact of corruption on FDI inflow is an ongoing
JFC
29,3
1042
Journalof Financial Crime
Vol.29 No. 3, 2022
pp. 1042-1063
© Emerald Publishing Limited
1359-0790
DOI 10.1108/JFC-05-2021-0106
The current issue and full text archive of this journal is available on Emerald Insight at:
https://www.emerald.com/insight/1359-0790.htm
academic and policy debate withoutconsensus. On the one hand, corruption is suggested to
grease the wheels of public administration, and hence, beneficial (Yang et al.,2018;
Ledyaeva et al., 2015). In contrast, an alternative proposition suggests that it sands the
wheels of public administration, leading to suboptimal economic gains (Ajide and Raheem,
2016;Kurul and Yalta, 2017;Kasasbeh et al.,2018). For instance, whereas Zander (2021)
argues that corruption facilitates the valuable expansion of the shadow economy during
critical periods such as recession, Habib and Zurawicki (2002a) contend that corruption
harms FDI due to the effect on operational efficiency. Other studies have found no
significant impact of corruption or the control thereof, on FDI (Okafor, 2015;Hoa and Lin,
2016;Goswami and Haider,2014).
Amid this confusion and in an apparent effort to solve this ambiguity,new studies have
attempted to unpack variouscorruption and/or FDI dimensions to assess the impact thereof.
This approach is premised on the argumentthat corruption and FDI are dynamic and multi-
dimensional variables, with mutually exclusive facets and independent variability. As an
example, Hakkala et al. (2008) differentiates between horizontal and vertical FDI and find
evidence that whereas horizontal FDI is constrained by corruption, vertical FDI is
stimulated by corruption. Similarly, Godinez and Liu (2015) confirm that FDI reacts
differently based on the relative depth of corruption in the host versus home country.
Specifically, higher levels of corruption in the host country relative to the home country
adversely impact FDI inflow to the host country. Luu et al. (2019) also confirm that the
impact of corruption is different based on whether FDI is a cross-border merger and
acquisition (M&A) or a greenfield investment.Specifically, M&A is impeded by corruption,
while greenfieldFDI is facilitated by corruption.
Extant empirical evidence in recent times has been compelling in confirming the
relevance of contextual considerations of systematic and behavioural phenomena (such as
corruption), arguing that such systematic rigidities and behavioural mechanisms do not
replicate, in whole, across several nations. In that regard, it is challenging to extract any
sound generalized conclusions from the plethora of aggregated studies that often assume,
without scientific cause, a linearrelationship of various variables on FDI. Economic models
based on aggregated data from multiplenations assume substantive structural, cultural and
systematic homogeneity between economic and geographical blocs [1]. However, the
empirical evidence for such sweeping assumptions is thin. Instead,the evidence confirms a
contextual distinctiveness of the impact of various variables on FDI primarily based on
national characteristics and other geopolitical considerations (Hakkala et al.,2008;Godinez
and Liu, 2015;Luu et al.,2019;Zangina and Hassan, 2020). It is conceivable that the varied
but related components and dynamic nature of FDI, including its medium to long term
focus, will affect short-run versuslong-run reaction to corruption. Based on this conception,
FDI may also react differently to a rise or reduction of corruption (Zangina and Hassan,
2020).
Even if we accept the argument of homogeneity in the structure of economies and
national characteristics adopted by studies that have considered Africa (or various sub-
regional blocs within Africa) as substantively homogenouseconomic and political systems,
we cannot assume substantive homogeneity in cultural, social and behavioural attributes.
The incidence of corruption can partially be attributed to cultural, social, behavioural and
systematic characteristics (Stathopoulou et al.,2021). If this view is accepted, juxtaposed
against the evidence that African economiesare widely diversified across various economic,
institutional and geographicalmatrixes, then a compelling case is made to unpack different
studies that use aggregated data sets. Impliedly, an alternative view is to focus on the
country-by-country impact of corruption on FDI. For example, Ghana is only one of four
ADRL
estimation of
corruption
1043
To continue reading
Request your trial