Content uploaded by Jacob Augustine
Author content
All content in this area was uploaded by Jacob Augustine on Jun 29, 2023
Content may be subject to copyright.
5
Feedback or no Feedback: Understanding the Interaction between Foreign Direct
Investment and Economic Growth in Nigeria
Okon J. Umoh1, Augustine O. Jacob2 and Chuku A. Chuku3
Abstract
This study sets out to investigate the relationship between foreign direct investment and economic growth in
Nigeria, between 1970 and 2008. Given the peculiar resource-based structure of the Nigerian economy,
the study also focused on identifying the key determinants of FDI flows to Nigeria. The methodology
adopted to carry out the study was influenced by prevalent economic theory which postulates that foreign
direct investments and economic growth may be jointly determined in an economy. In other words, there is
positive feedback between FDI and growth. Consequently, this study adopted simultaneous equation
models and single equation models to study the relationship between economic growth and foreign direct
investments and to identify the determinants of foreign direct investment in Nigeria. Specifically, the three-
stage and two-stage least squares approach and the error correction models were implemented respectively
to investigate these relationships. The results obtained show that FDI and economic growth are jointly
determined in Nigeria, and there is positive feedback from FDI to growth and from growth to FDI in
Nigeria. Furthermore, the results identified four core determinants of FDI in Nigeria: openness, global
rate of return on investment, government consumption and the exchange rate. The overall insight from the
study is that economic growth can be further enhanced with more FDI, therefore the recommendation that
emerged from the study is that relevant authorities should intensify efforts at attracting more foreign direct
investments.
Keywords: FDI, simultaneous equation systems, economic growth, endogeneity, Nigeria
JEL Classification: F21, E22, O47
INTRODUCTION
Given Nigeria‟s enormous natural resource base and large market size, Nigeria qualifies
to be a major recipient of FDI in Africa. Unfortunately, the level of FDI attracted by
Nigeria is still sub-optimal compared with its resource base and potential need. Recent
studies have shown that foreign direct investment (FDI) is what is needed to bridge the
1 Senior Lecturer, Department of Economics, University of Uyo, Uyo, Nigeria
2 Rector, Uyo City Polytechnic, Akwa Ibom, Nigeria
3Corresponding author, Department of Economics, University of Uyo, Uyo Nigeria. email
chukuachuku@gmail.com; chukuchuku.econs.lect@uniuyo.edu.ng phone: +2348067247177
Vol. 11, No.2 Journal of Monetary and Economic Integration
6
savings-investments gap that exists in Africa in general and Nigeria in particular. Prior to
the 1970s, Foreign Direct Investment (FDI) was not seen as an instrument of economic
development. The perception of FDI as parasitic and retarding the development of
domestic industries for export promotion had engendered hostility to multi-national
companies and their direct investments in many countries. However, the consensus now
is that FDI is an engine of growth as it provides the much needed capital for investment,
increases competition in the host country industries, and aids local firms to become more
productive by adopting more efficient technologies or by investing in human and/or
physical capital and it is more stable than other forms of capital flows (Ajayi, 2006).
As a result of the potential role of foreign direct investment in accelerating growth and
economic transformation, many developing countries and Nigeria in particular seek such
investments to accelerate its development efforts. Attracting FDI has therefore become a
major component of development strategy in Nigeria, particularly because it will help to
bridge the savings-investment gap, but also because it can assist in the attainment of
vision 2020:20 targets. Indeed, in 2011, the Federal government of Nigeria set up the
Ministry of Trade and Investment, which is primarily concerned with facilitating trade
and foreign and domestic investments in the economy. If the theory says that FDI
should cause growth in an economy, then the next line of thinking should be to
empirically verify if these claims are valid for a resource rich developing economy like
Nigeria.
The proposition made in this paper is that FDI facilitates economic growth on the one
hand, and on the other hand, economic growth attracts foreign direct investments into
Nigeria. In other words, FDI and economic growth are both endogenously determined
and have positive feedback effects in Nigeria. Hence, the objectives of the paper are as
follows: (i) to examine the relationship between FDI inflows and economic growth in
Nigeria with a view to uncovering if there are any feedbacks in the relationship, (ii) to
indentify the core determinants of FDI inflows into Nigeria and (iii) to analyse the
stylized facts about FDI and growth in Nigeria. The study is justified particularly because
it does not simply assume endogeniety, but actively tests for endogeniety of FDI and
economic growth in Nigeria. Also, it differs from all other studies in scope (1970-2008).
This gives the study an edge because it examines the FDI-growth relation in the near-
contemporary context, taking account of past trends and recent developments in the
global financial market.
The rest of the paper is as follows. Section 2 presents a comprehensive review of the
theoretical and empirical literature on the FDI-growth relation. Section 3 contains a
description of the methodology employed in the study. In Section 4, the results from the
estimations are presented along with their theoretical discussion. Finally, Section 5
contains the policy implications and discussion.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
7
THEORETICAL AND EMPIRICAL LITERATURE
THEORIES OF FOREIGN DIRECT INVESTMENT
The Eclectic or Ownership, Location and Internalization (OLI) Theory of FDI
The most robust and comprehensive economic theory of FDI is the eclectic or OLI
framework developed by John Dunning in the late 1970‟s (See Dunning, 1979). This
theory attempts to explain the existence, activities and strategies of multinational
enterprises (MNEs) through the synthesis of macro- and micro-economic determinants
of FDI flows.
In this way, Dunning‟s theory of international capital flows integrates industrial
economics and location theories within the broader framework to international trade and
investments. The eclectic framework identifies three sources of advantages that are
preconditions for firms to engage in international production i.e., to become MNEs: (1)
Ownership (O) advantage; (2) Location (L) advantage and (3) Internalization (I)
advantage.
Ownership advantage refers to the need for MNEs to possess firm-specific competitive
advantages over domestic firms in serving particular markets. These advantages may
include both tangible and intangible sources of advantages and arise from the monopoly
control of these assets by the MNEs, often reflecting the factor endowments and
characteristics of their home countries. The O-advantage offers the potential for
substantial increasing returns to scale, resulting from the relatively low or zero marginal
cost incurred in transferring them across international borders. Location advantage arises
from the profitable combination of MNEs ownership advantages with inputs,
intermediate output and/or services originating from outside their home country, i.e.
through international production. Location advantages therefore provide MNEs with an
incentive to locate at least some part of their activities in another (host) country rather
than at home. MNE location decisions are therefore founded upon the actual and
perceived competitive advantages of potential host countries.
The market imperfections paradigm
This theory was developed by Kindleberger (1969). Here, FDI is approached from the
standpoint of the perfectly competitive model of neoclassical economics by asserting that
in a world of pure competition, direct investments will not exist. That is, when all
markets operate efficiently; when there are no external economies of production or
marketing; when information is costless and there are no barriers to trade or
competition; international trade is the only possible form of international involvement
(Calvet, 1980). Logically, it follows that it is the departure from the model of perfect
competition that must provide the rationale for foreign direct investment. According to
Hymer (1970) who originally developed this theory, two conditions have to be fulfilled to
Vol. 11, No.2 Journal of Monetary and Economic Integration
8
explain the existence of direct investment (1) foreign firms must possesses countervailing
advantages over the local firms to make such investment viable, and (2) the market for
the sale of this advantage must be imperfect. It was these two conditions that framed
Kindleberger‟s mind in coming up with the market imperfectionist theory of FDI.
Specifically, he came up with the following taxonomy: imperfections in goods markets,
scale economies and government imposed distortion (Kindleberger, 1969). This
taxonomy is the basis for all the other theories of FDI that are discussed in this chapter.
Hence, theories of FDI can be classified along four broad categories: (1) market
disequilibrium hypothesis; (2) government imposed restriction (3) market structure
imperfections and (4) market failure imperfections.
The common feature found in all hypotheses in group (1) will be the transitory nature of
foreign direct investment. FDI is an equilibrating force among segmented markets which
eventually comes to an end when equilibrium is re-established; that is, when rates of
return are equalized among countries. The unifying characteristics in Group (2) will be
the role played by either host or home government in providing the incentives to invest
abroad. Group (3) will include theories in which the behaviour of the firm deviates from
normal conditions under perfect competition, through their ability to influence market
prices. Finally, Group (4) will contain classified theories which depart from the
technical assumptions behind the model of perfect markets; that is, the assumptions
about production techniques and commodity properties. This last category will deal
basically with those phenomena which lead to market failure, or cases where “the
decentralization efficiency of that regime of signals, rules and built-in-sanctions which
defines a price-market system” will fail (Bator, 1958; Calvet, 1980). Several factors
support this classification scheme. In the first place, the order in which the categories
appear corresponds roughly to the chronological order in which new explanations of the
FDI phenomenon have occurred. They range from the old view of an integrated
approach to foreign direct investment and portfolio capital flows among countries, to
more recent versions of FDI as a spin-off of welfare economics.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
9
Market disequilibrium hypothesis of FDI
Following the market disequilibrium theory of FDI, the notion of a perfect economy and
perfect market requires that prices everywhere are adjusted to bring supply and demand
into equilibrium. However, this condition does not hold because of segmentation in the
world market, and unequal rates of return from various countries. Under these
conditions, flows of FDI would take place until markets returned to stability. Instances
of disequilibrium conditions that provide incentives to invest abroad are numerous. They
basically apply to factor markets and foreign exchange markets (Calvet, 1980). Currency
overvaluation is perhaps the most salient example of these disequilibrium hypotheses. A
currency may be defined as overvalued when at the prevailing rates of exchange;
production costs for tradable goods in the country are on the average, higher than in
other countries (Ragazzi, 1973). Such an occurrence creates an opportunity for profit
making by holding assets in undervalued currencies with the expectation that, once
equilibrium in the foreign exchange market is re-established, capital gains will be realized.
In the meantime, there is an incentive to locate production of internationally traded
commodities in countries with undervalued currencies and to purchase income
producing assets with overvalued money. The cardinal point about this hypothesis is that
once exchange rates return to equilibrium, the flow of FDI should stop.
Another aspect of the disequilibrium hypothesis deals with the relative rates of profits in
different markets. This is basically referred to as the capital market disequilibrium
hypothesis (see Calvet, 1980). It implies that for a given level of risk, rates of return are
not equalized internationally by portfolio capital flows, due to inefficiencies in security
markets such as thinness or lack of disclosure. Therefore, the only way that rates of
return on real assets can be brought to equilibrium is by flows of direct investments. The
process is however self-destructive because firms from low yield countries will invest in
countries with high yields until rates of return are brought to equilibrium, then the
foreign direct investment will cease. Finally, this equilibrium situation may arise in
technology markets. Rates of technical and technological innovation may vary among
nations, thereby placing some countries in leadership positions with respect to new
products and processes. The origins of superior knowledge can be traced back to
superior R&D performance or merely chance factors- such as, a breakthrough in
scientific knowledge. In any event, firms in countries where technology is relatively
advanced would find profitable opportunities abroad and would therefore have an
incentive to invest overseas.
Oligopolistic framework
The previous causes of disequilibrium hypotheses and government-induced distortions
are compatible with a “relatively” competitive market structure. Market structure
imperfections, on the other hand, refer to deviations from purely market-determined
prices brought about by the existence of monopolistic or oligopolistic market
characteristics. In this perspective, foreign direct investment becomes the outgrowth of
Vol. 11, No.2 Journal of Monetary and Economic Integration
10
industrial organization. The recognition that foreign direct investment belongs to the
realm of industrial organization goes back to Hymer‟s writings in 1976. Since then, it has
received much attention and has become the most popular approach to date. (Bergsten,
et al, 1978). There are two essential characteristic which set oligopolistic industries apart
from competitive ones. Not all barriers to entry lend themselves to direct expansion
abroad. Caves (1971) considered product differentiation in the home market as being the
critical element giving rise to foreign investment. The successful firm, producing a
differentiated product, controls knowledge about servicing the domestic market that can
be used at little or no cost in other national markets. This provides the motivation for
investing abroad, as long as the means to protect the product exist; such as patents and
copyrights.
Other contributions to the oligopolistic feature of direct investment include models
which explicitly take into account the interdependence of firms in the industry. The most
publicized is, perhaps, the product life cycle hypothesis (Vernon, 1966) where firms react
to the threat of losing marketsas the product matures, by expanding overseas and
capturing the remaining rent from the product‟s development. Variations on this
approach include the “follow-the-leader” case, where the investment moves of one firm
trigger similar moves by other leading firms in the industry (Knickerbocker 1974). The
other is “the exchange of threats” hypothesis, where oligopolistic firms compete with
each other by establishing subsidiaries in each other‟s markets (Graham, 1974). Finally,
and particularly in Caves‟s (1971) argument, the static concept of product differentiation
is related to the notion of intangible capital in the form of knowledge, yet the two do not
imply the same form of international involvement. Indeed, industries with an unchanging
product mix since equality changes, brand changes, etc. pose no difficulties for
arguments (Baumann 1975). However, as will be seen in the next section, the transfer of
intangible capital in the form of knowledge does pose serious problems.
DETERMINANTS OF FDI
Over time, a number of studies have been carried out to examine/analyse the various
determinants of FDI in Africa and in Nigeria specifically. In one or two cases, Africa is
shown to be different from the rest of the world in terms of the various factors affecting
foreign direct investment. According to Asiedu (2002), policies that have been successful
in other regions may not be equally successful in Africa. The second is that economic
policy does not matter for FDI. The findings of the few studies on the determinants of
FDI in Africa and developing countries have been contradictory in many cases. There is
a dearth of empirical work that is solely concentrated on the determinants of FDI in
African countries or Nigeria. In most of the studies that have been carried out, only a
limited number of African countries are included. For example, Gastanaga, et al. (1998)
in studying the impact of FDI on economic growth, considered a total of 49 countries,
only 6 of which are in sub-Saharan Africa (SSA), while Schneider and Frey (1985)
considered 51 countries, of which 13 are in SSA.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
11
In their econometric analysis of the determinants of FDI using panel data, Elbadawi and
Mwega (1997) argue that while market size is relatively unimportant in explaining FDI
flows to Africa, economic growth is an important determinant. They found that a
depreciation of the real effective exchange rate, an increase in a country‟s openness to
trade and the expansionary effects of fiscal balance have positive impacts on FDI. It is
also shown that an improvement in removing restrictions and providing good conditions
for private initiative have important bearings on FDI inflows, while the number of
political upheavals has a negative bearing. Terms of trade shocks and the level of
schooling are found to have little impact on FDI into Africa. Incidents of war and
African regional integration arrangements are found to have limited impacts on FDI
flows. Two recent studies also concentrate on Africa. The first, by Schoeman, et al.
(2000), which is limited to South Africa, analysed how government policy (mainly deficit
and taxes) affects FDI. The second set of papers is by Asiedu (2002, 2004). Using cross-
section data on 71 developing countries, Asiedu (2002) attempted to answer the
following set of questions: What factors drive FDI to developing countries? Are these
factors equally relevant for FDI to SSA? Why has SSA attracted so little FDI? Why has
SSA been relatively unsuccessful in attracting FDI despite policy reform? Is Africa
different? The analysis is focused on only three main variables – the return on
investment, availability of infrastructure and openness to trade – and does not take into
account natural resource availability, which is an important determinant of FDI to Africa.
Asiedu (2004) concludes that: Countries in SSA have on average received less FDI than
countries in other regions by virtue of their geographical location. Both higher return on
investment and better infrastructure have positive impact on FDI to non-SSA countries,
but no impact on FDI to SSA; openness to trade promotes FDI to SSA and non-SSA
countries; the marginal benefit from increased openness is less for SSA, suggesting that
trade liberalization will generate more FDI to non-SSA countries than SSA countries.
Her results imply that Africa is different and that factors attracting FDI to other regions
may not be equally applicable in Africa. This implies that the success stories in other
places cannot in some cases be replicated in Africa. Three policy implications arise from
the results of the empirical work. (1). African countries need to liberalize their trade
regime in order to enhance FDI flows. The full benefit of trade liberalization is only
achievable if investors perceive the reform not only credible but irreversible. (2) Policies
that have worked in other countries cannot be blindly replicated in Africa, since these
policies may have different impacts on Africa. (3) Africa is overly perceived as risky.
Consequently, countries in the region receive less FDI by virtue of their geographical
location. To dispel the myth, there is need to disseminate information about the
continent.
In another paper, Asiedu (2003) used panel data on 22 African countries for the period
1984–2000 to examine empirically the impact of several variables including natural
Vol. 11, No.2 Journal of Monetary and Economic Integration
12
resource endowment, macroeconomic instability, FDI regulatory framework, corruption,
effectiveness of the legal system and political instability on FDI flows. The paper
debunks the notion that FDI in Africa is solely driven by natural resource availability and
concludes that natural resource endowment, large markets, good infrastructure and an
efficient legal framework promote FDI, while macroeconomic instability, corruption,
political instability and investment restrictions deter investment flows. These results
imply that African governments can play major roles in promoting FDI to the region
through appropriate policy framework.
In the short and medium term, government can increase their FDI by streamlining their
investment regulation framework, implementing policies that promote macroeconomic
stability and improving infrastructure. In the long run, more FDI can be achieved by
curbing corruption, developing a more efficient legal framework and reducing political
instability (Asiedu, 2003). Morisset (2000) focuses exclusively on Africa and controls for
natural resource availability. He identified which African countries have been able to
attract FDI by improving their business climate. Evidence from the countries shows that
proactive policies and reoriented governments can generate FDI interest. He makes the
point that by implementing policy reforms, African countries can also be successful in
attracting FDI that is not based on natural resources or aimed at the local market, but
rather on regional and global markets. Using panel data for 29 countries over the period
1990–1997 Morisset (2000) finds that GDP growth rate and trade openness have been
positively and significantly correlated with the investment climate in Africa. On the other
hand, the illiteracy rate, the number of telephone lines per capita and the share of the
urban population (a measure of agglomeration) are major determinants in the business
climate for FDI in the region. Political and financial risk as measured by the International
Country Risk Guide (ICRG) and the International Investors ratings did not appear
significant in his regression.
One major deterrent to FDI flows cited in the literature is uncertainty. Uncertainty is also
a known factor plaguing Africa‟s development strategy. Empirical studies of the
relationships between FDI and uncertainty in developing countries are very few. Two
studies, by Ramasamy (1999) for Malaysia and Lehmann (1999) for developing countries,
found a negative relationship between uncertainty and FDI in developing countries.
Even fewer studies address the connection between uncertainty and FDI in Africa. While
Bennell (1995) and Pigato (2001) highlighted the role of uncertainty, none of them
formally addresses the impact of both economic and political uncertainty in African
countries.
Pigato (2001) examined how uncertainty affects FDI flows to African economies by
analysing FDI flows from the United States, US manufacturing FDI and US non-
manufacturing FDI flow to sampled host countries in Africa. Using a generalized
autoregressive heteroscedastic model, the study concludes that the impact of uncertainty
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
13
on the flow of FDI from all sources is insignificant; that in the aggregate with regard to
FDI from the United States, economic and political uncertainties are not major concerns;
that for US manufacturing FDI, only political instability and government policy
commitment are important factors, whereas for US non-manufacturing FDI, economic
uncertainties are the major impediments only when coupled with political instability and
debt burden of host countries, and finally, that other economic factors such as labour,
trade connections, size of the export sector, external debt and market size are also
significant in affecting FDI flow to Africa.
THE RELATIONSHIP BETWEEN FDI AND ECONOMIC GROWTH
The impact of FDI on growth is manifold. Although, there is yet no consensus on the
relationship between FDI and growth, there is a growing convergence in recent times
that FDI is positively correlated with growth. This view has derived its theoretical
underpinning from recent developments in growth theory which highlights the
importance of improved technology, efficiency and productivity in stimulating growth
(Lim, 2001). In this regard, FDI‟s contribution to growth comes through its role as a
conduit for transferring advanced technology from the industrialized to the developing
economies. For instance, Findlay (1978) postulated that FDI increases the rate of
technical progress in the host country through a “contagion” effect from the more
advanced technology and management practices used by foreign firms.
The contagion or knowledge diffusion effect often referred to as “externalities” or
“efficiency spillovers” can lead to improvements in productivity and efficiency in local
firms in several ways. In a simple form, a spillover can occur when a local firm improves
its productivity by copying some technology used by corporation/affiliate MNC in the
local market. Another aspect of the contagion effect occurs when local firms are forced
to use existing technology and resources more efficiently, or to search for more efficient
technologies, because an MNC‟s entry has increased competitive pressures in the local
market. Further, spillovers can occur when an affiliate demonstrates new techniques to
and trains local workers, who later accept employment in local firms or start their own
firms (Lim, 2001).
Borensztiein, et al. (1998) asserted that local firms have an opportunity to improve their
efficiency by learning and interacting with foreign firms. Through this medium, FDI can
raise the quality of domestic human capital and improve the know-how, and managerial
skills of local firms. This is usually referred to the learning-by-doing effect (Li and Liu,
2005; Bende-Nabende, Ford and Sen, 2003). Interestingly, Hermes and Lensink (2000)
summarized the different channels through which positive externalities associated with
FDI can occur, namely: (i) the competition channel, where increased competition leads
to increased productivity, efficiency and investments in human and/or physical capital.
Increased competition may lead to changes in the industrial structure towards more
competitiveness and more export-oriented activities (ii) training channel through
Vol. 11, No.2 Journal of Monetary and Economic Integration
14
increased training of labour and management. (iii) linkage-channel, whereby foreign
investment is often accompanied by technology transfer. Such transfers may take place
through transactions with foreign firms and (iv) domestic firms imitate the more
advanced technologies used by foreign firms, commonly termed as the “demonstration
channel”, Romer (2001) looks at technology as a non-rival input and at foreign direct
investment as a source of technological advance. In this case, the FDI effect is
unequivocally positive. On the other hand, Balasubramanyan, Salisu and Sapsford (1996)
suggests that the growth effect of FDI might be positive for export promoting (EP)
countries, but negative for import substituting (IS) ones, the reduction of foreign import
goods in the domestic market reduces competition and efforts to improve efficiency
among the domestic firms. Reis (2001) used an endogenous growth model to evaluate
the growth effects of FDI when the investing firm‟s profits may be repatriated. She
found that in equilibrium, foreign firms replace all domestic firms in the R&D sector. In
this model, FDI only adds a positive effect to growth if the world interest rate is lower
than the home interest rate.
Li and Liu (2005) found that FDI not only affects growth directly, but also indirectly
through its interaction with human capital. Further, they find a negative coefficient for
FDI when it is regressed with the technology gap between the source and host economy
using a large sample, Borensztein et al. (1998) found similar results i.e. that inward FDI
has positive effects on growth with the strongest impact, coming through the interaction
between FDI and human capital. De Mello (1999) found positive effects of FDI on
economic growth in both developing and developed countries, but concludes that the
long-run growth in host countries is determined by the spillovers of knowledge and
technology from investing countries to host countries. Similarly, Balasubramanyan, Salisu
and Sapsford (1996) found support for their hypotheses that the growth effect of FDI is
positive for export promoting countries and potentially negative for import-substituting
ones.
Alfaro et al. (2004) and Durham (2004) focused on the ways in which the FDI effect
depends on the strength of the domestic financial markets of the host country. They
both found that only countries with well developed banking and financial systems benefit
from FDI. In addition, Durham (2004) found that only countries with strong
institutional and investor-friendly legal environments are likely to benefit from FDI
inflows. In another work, Hsiao and Shen (2003) add that a high level of urbanization is
also conducive to a positive impact of FDI on growth. Comparing evidence from
developed and developing countries, Blonigen and Wang (2005) argued that mixing
wealthy and poor countries is inappropriate in FDI studies. They note that the factors
that affect FDI flows are different across the income groups. Interestingly, they find
evidence of beneficial FDI only for developing countries, and not for the developed
ones, while they find the crowding-out effect of FDI on domestic investment to hold for
the wealthy group of nations.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
15
Recently, Vu and Noy (2009) carried out a sectoral analysis of foreign direct investment
and growth in developed countries. They focused on the sector specific impacts of FDI
on growth. They found that FDI has positive and no statistically discernible effects on
economic growth through its interaction with labour. Moreover, they found that the
effects seem to be very different across countries and economic sectors. Carkovic and
Levine (2005) argue that the positive results found in the empirical literature are due to
biased estimation methodology. When they employed a different estimation techniques
i.e. Arellano-Bond generalized moment of methods (GMM), they found no robust
relationship between FDI inflows and domestic growth. In line with the notion that
there is an endogenous relationship between FDI and economic growth, Ruxanda and
Muraru (2010) investigated the relationship between FDI and economic growth in the
Romanian economy, using simultaneous equation models. They obtained evidence of the
bi-directional connection between FDI and economic growth, meaning that incoming
FDI stimulates economic growth and in its turn, a higher GDP attracts FDI. In a paper
most similar to this work, Li and Liu (2005) investigated the relationship between FDI
and economic growth based on a panel of 84 countries, using both single equation and
simultaneous equation systems. They found that FDI affects growth indirectly through
its impact on human capital. This work is similar to their own in that we use both single
equation and simultaneous equation systems. However, our work is different in that it is
country specific (Nigeria) and involves a longer time frame (1970-2008).
The consensus in the literature appears to be that FDI spillovers depend on the host
country‟s capacity to absorb the foreign technology and the type of investment climate
(Obwona, 2004). The review here and in the references provided, shows that the debate
on the impact of FDI on economic growth is far from being conclusive. The role of FDI
seems to be country specific, and can be positive, negative or insignificant, depending on
the economic, institutional and technological conditions in the recipient countries. Most
studies on FDI and growth are cross-country evidences, while the role of FDI in
economic growth can be country specific. Further, only a few of the country specific
studies actually took conscious note of the endogenous nature of the relationship
between FDI and growth in their analyses, thereby raising some questions on the
robustness of their findings. Finally, the relationship between FDI and growth is
conditional on the macroeconomic dispensation the country in question is passing
through. In fact, Zhang (2001) asserts that “the extent to which FDI contributes to
growth depends on the economic and social condition or in short, the quality of the
environment of the recipient country”. In essence, the impact FDI has on the growth of
any economy may be country and period specific, and as such there is the need for
country specific studies. This discovery from the literature is what provides the
motivation for this study on the relationship between FDI and economic growth in
Nigeria.
Vol. 11, No.2 Journal of Monetary and Economic Integration
16
THE FDI-GROWTH RELATION IN NIGERIA
There are several Nigeria-specific studies on the relationship between FDI and economic
growth. Some of the pioneering works include Aluko (1961), Brown (1962) and Obinna
(1983). These authors separately reported that there is a positive linkage between FDI
and economic growth in Nigeria. Edozien (1968) discussed the linkage effect of FDI on
the Nigerian economy and submits that these have not been considerable and that the
broad linkage effects were lower than the Chenery-Watanabe average. Oseghale and
Amonkhienan (1987) found that FDI is positively associated with GDP, concluding that
greater inflows of FDI will spell a better economic performance for the country. Odozi
(1995) placed special emphasis on the factors affecting FDI flows into Nigeria in both
pre and post structural adjustment programme (SAP) eras and found that the macro
policies in place before SAP where discouraging investors. This policy environment led
to the proliferation and growth of parallel markets and sustained capital flight.
Adelegan (2000) explored the seemingly unrelated regression model (SUR) to examine
the impact of FDI on economic growth in Nigeria and found that FDI is pro-
consumption, pro-import and negatively related to gross domestic investment. In
another paper, Ekpo (1995) reported that political regime, real income per capita,
inflation rate, world interest rate, credit rating and debt service were the key factors
explaining the variability of FDI inflows into Nigeria. Similarly, Ayanwale and Bamire
(2001) assessed the influence of FDI on firm level productivity in Nigeria and reported
positive spillover of foreign firms on domestic firm productivity. Ariyo (1998) studied
the investment trend and its impact on Nigeria‟s economic growth over the years. He
found that only private domestic investment consistently contributed to raising GDP
growth rates during the period considered (1970–1995). Furthermore, there is no reliable
evidence that all the investment variables included in his analysis have any perceptible
influence on economic growth. He therefore suggested the need for an institutional
rearrangement that recognizes and protects the interest of major partners in the
development of the economy
A common weakness that has been identified in most of these studies is that they failed
to control for the fact that most of the FDI inflows to Nigeria has been concentrated on
the extractive industry (to oil and natural resources sector). Akinlo (2004) specifically
controlled for the oil, non-oil FDI dichotomy in Nigeria. He investigated the impact of
foreign direct investment (FDI) on economic growth in Nigeria, using an error
correction model (ECM). He found that both private capital and lagged foreign capital
have small, and not a statistically significant effect on economic growth. Further, his
results support the argument that extractive FDI might not be growth enhancing as
much as manufacturing FDI. Examining the contributions of foreign capital to the
prosperity or poverty of LDCs, Oyinlola (1995) conceptualized foreign capital to include
foreign loans, direct foreign investments and export earnings. Using Chenery and Stout‟s
two-gap model (Chenery and Stout, 1966), he concluded that FDI has a negative effect
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
17
on economic development in Nigeria. Further, on the basis of time series data, Ekpo
(1995) reported that political regime, real income per capita, rate of inflation, world
interest rate, credit rating and debt service were the key factors explaining the variability
of FDI into Nigeria.
Anyanwu (1998) paid particular emphasis on the determinants of FDI inflows to Nigeria.
He identified change in domestic investment, change in domestic output or market size,
indigenization policy, and change in openness of the economy as major determinants of
FDI inflows into Nigeria and that effort must be made to raise the nation‟s economic
growth so as to be able to attract more FDI. Ayanwale (2007) investigated the empirical
relationship between non-extractive FDI and economic growth in Nigeria and also
examined the determinants of FDI inflows into the Nigerian economy. He used both
single-equation and simultaneous equation models to examine the relationship. His
results suggest that the determinants of FDI in Nigeria are market size, infrastructure
development and stable macroeconomic policy. Openness to trade and human capital
were found not to be FDI inducing. Also, he found a positive link between FDI and
growth in Nigeria. Our work is similar to that of Ayanwale (2007), in that we seek to
examine the determinants and impact of FDI on growth in the Nigerian economy.
However, our work is improved because we consider a longer time frame (1970-2008),
whereas that of Ayanwale was (1970-2002) and we use a more robust system of equation
i.e. three stage least squares, 3SLS estimation methodology.
STYLIZED FACTS ON FDI AND GROWTH IN NIGERIA
The stylized facts about the details of FDI inflow into Nigeria are shown for the period
1970 to 2008 in Table 1. FDI inflows for the period under review ranged from N128.6
million in 1970 to N2,569,509million in 2008. The average FDI inflow to Nigeria has
been N19,186 million and the total FDI inflows to Nigeria between 1970 to 2008 is
N7,690,254.3million. These values are significant compared with FDI inflows to other
developing countries, and countries in Africa. The World Investment Report (2003)
ranked Nigeria as the third largest recipient of FDI in Africa.
Comparing the percentage of FDI to Nigeria‟s GDP, we observe that FDI forms a small
percentage of Gross Domestic Product in Nigeria. FDI‟s contribution to GDP was
particularly low between the years 1973 to 1986, as it fluctuated between 0.3 and 0.9% of
GDP. FDI contributed most to GDP in 2007, as the ratio was highest that year at 9.27.
Within the period under review, FDI has on average contributed 2.01% to GDP. This is
an indication that FDI is a driver of growth in Nigeria, and if properly harnessed, the
positive feedback from FDI to GDP can be maximized.
Vol. 11, No.2 Journal of Monetary and Economic Integration
18
Table 1: Foreign Direct Investment in Nigeria, 1970-2008
YEAR FDI(N million) GDP(N million) FDI AS % OF GDP
1970 128.6 6650.9 1.933573
1975 253 26655.78 0.949137
1980 -404.1 47619.66 -0.848599
1985 434.1 69146.99 0.627793
1990 4686 312139.74 1.501251
1995 75940.6 2702719.13 2.809785
2000 115952.2 4725086 2.45397
2005 341717.2 18564594.7 1.840693
2006 997860.4 23280715 4.28621
2007 2229660.8 24048480 9.271525
2008 2569509 2876545.5 8.32621
Source: compiled from CBN Statistical Bulletin (2009)
Table 2 displays some key indicators of FDI in Nigeria as reported by the United Nations Conference on Trade
and Development (UNCTAD). We observe that FDI inflows have largely been more than outflows except for the
year 1990, when FDI inflows was $588 million as against a large outflow of $1,824 in the same year. FDI
inflows to Nigeria peaked in 2005 at $2,751 million. This may have reflected the relative political stability at that
time and the numerous macroeconomic incentives that were instituted during the same period.
Source: UNCTAD Database www.unctad.org/statistics (accessed October 2, 2010)
Table 2: FDI indicators for Nigeria (1985-2005)
Indictors 1985 1990 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005
FDI Inflows (US$
million) 486 588 1079 1593 1539 1051 1005 930 1104 1281 2435.5 2373 2751
FDI Outflows
(US$ million) -15 1824 104 42 58 107 92 85 94 101 200.1 180.5 220.3
Inflows as % of
GCF 8 15 21 201 16 12 52 49 31 35 30.3 27.1 33.8
Outflows as % of
GCF 0 47 2 1 1 1 5 52 3 3 2.5 3 4.4
FDI Inward stock
(US$ million) 4417 8072 14065 15658 17198 18249 19254 20184 21289 22570 23564 23889 24194
FDI Outward
stock (US$
million) - 2586 3975 4017 4017 4182 4273 43582 4452 4553 47652 4928 4589
Inward stock as
% of GDP 16 28 50 44 47 57 55 42 42 42 44 57 41
Outward stock as
% of GDP - 9 14 11 11 13 12 9 9 9 10 8.5 9
19
FDI inflows, as a percentage of gross capital formation (GCF) have performed fairly
well. Though its counterpart on the percentage of FDI outflows to gross capital
formation has been very weak during the period under review, it has however been
growing. The percentage of FDI inflows to gross capital formation (GCF) peaked in
1996 at 201% and has never attained that level again. The trend has however not been
consistent as its percentage has been increasing and decreasing intermittently. As at 2005,
the ratio of FDI to GCF was 33.8%. FDI outflows as a percentage of gross capital
formation has relatively been insignificant as it assumed the value of 1% between 1993
and 1998.
On the ratio of FDI inflows and outflows to GDP, we observe that FDI has significantly
contributed to gross domestic product in Nigeria. For example in 1998, 1999 and 2004,
FDI contributed 55%, 57% and 57% respectively to GDP in Nigeria. This goes a long
way to suggest that the Nigerian economy has been relatively open and that FDI is a key
driver of economic activities in Nigeria. FDI outflows as a percentage of GDP has
fluctuated between 9 and 14 percent between 1985 and 2005, suggesting also that a
significant proportion of Nigerian‟s resources constitute direct investment in other
countries.
METHODOLOGY
ANALYTICAL FRAMEWORK
This study is primarily concerned with providing structural analysis for two key
relationships in the Nigerian case; (1) the growth impact of foreign direct investments,
(2) the determinants of foreign direct investment inflows to Nigeria. For the first
objective, the most appropriate analytical framework used, is the FDI-spillover model as
developed by Romer (1986). In the case of the determinants of foreign direct investment,
we follow Fedderke and Romm (2006) who developed a theoretical model of the
location of the investment activity as an explicit choice in an intertemporal context of
locating new capital stock either domestically or in an alternative foreign location.
SPILL-OVER ANALYTICAL FRAMEWORK FOR FDI AND GROWTH
FDI can be analytically linked to growth through a differentiated impact of FDI on
productivity of both domestic labour and domestic capital, through the transmission of
superior technology. The analytical structure, is therefore, in the spirit of Romer (1986).
The importance of FDI can be seen as closing the capital-gap identified by Romer (1993)
as the main obstacle facing developing countries trying to catch-up with advanced
countries. This gap is more on knowledge or human capital, than the gap in physical
capital. In the spirit of De Mello (1997); Ramirez (2000) and Fedderke and Romm
Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
Vol. 11, No.2 Journal of Monetary and Economic Integration
20
(2006), the analytical framework that links FDI to economic growth can be analyzed via
an augmented Cobb-Douglas production function, as follows:
=,,, =
1(1)
Where Y is real output, is the domestic capital, , is foreign capital, L is labour and
E refers to the externality or spillover effect (≠ 1) generated by the additions to the stock
of FDI. and are the shares of domestic labour and capital respectively, and A
captures the efficiency of production. Here, we assume that +< 1. For simplicity,
let the externality, E be represented by a Cobb-Douglas function of the type
= [,,
](2)
Where
denotes foreign owned capital. Combining Equations (1) and (2), we obtain
=+ 1+ 11 (3)
From Equation (2), ()()
=
, such that ≠ 0 implies that domestic
and foreign capital may either serve as substitutes or complements. This corresponds to
the crowding-out and crowding-in effects of FDI respectively (see Fedderke and Romm,
2006). Specifically, when ˃ 0, foreign direct investment crowds out domestic
investment (De Mello, 1997). By contrast, captures the spill over effect of foreign
direct investment on the productivity of capital and labour. It is therefore, possible to
interpret as the instantaneous or marginal effect of foreign capital on output, and as
the long-run or intertemporal elasticity of substitution between domestic and foreign
capital. Finally, we can generate the dynamic production function by taking the
logarithms and time derivatives of Eq. (3):
= + + 1 + 1
+ 1 . (4)
Where is the growth rate of i, and i stands for Y, A, L,, and respectively.
Portfolio Analytical Framework for Determinants of FDI
We follow the analytical framework put forth by Fedderke (2002) who proposed a model
of foreign direct investment by building on the portfolio theoretic framework with
intertemporal optimization. Under this approach, the core drivers of FDI fall into two
classes of determinants: (1) rates of return and (2) risk factors. The relationship is such
that FDI responds positively to rates of return and negatively to risk. We employ the
standard variation approach, and begin by defining the expected return on a portfolio of
capital assets faced by an agent which we denote as E(R), hence:
= + . . (5)
Where and are defined as the expected return on domestic and foreign capital
assets respectively. and are defined as cost of adjustment of domestic and foreign
capital assets holdings respectively. Cost of adjustment is held to arise due to information
and transactions costs associated with altering the composition of capital assets
portfolios.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
21
Returns on domestic assets are distinguished from returns on foreign assets by having a
non-zero probability of “expropriation‟ denoted by 0 1. Expropriation includes
factors such as the nationalization of assets, periods of domestic instability, capital
controls and direct or implicit taxes faced by foreign and domestic investors.
In line with Fedderke and Romm (2006), we postulate that
= 21 , 0 1, α,β> 0 6
= 2, ,> 0 (7)
Where , denotes domestic and foreign capital holdings , respectively. For
adjustment costs, we assume that the cost of adjustment is increasing in the magnitude of
adjustment for both domestic and foreign capital assets. Thus we have:
= ′+ ′2, ,> 0 (8)
= ′+ ′2, ,> 0 (9)
Note that variation in the adjustment costs of domestic capital asset holdings is perhaps
the prime policy handle available to domestic policy makers, together with the ability to
change expropriation risk. All of , might be affected by policy interventions
that raises the friction costs of moving capital across international boundaries (Feddeke
and Romm, 2006). The net present value of the expected return on a portfolio of capital
assets over an infinite time horizon is then:
,=
∞
0
(10)
Details about the Euler solution derived from Eq.(10), the optimal time path for
investment in domestic assets and the optimal time path for investments in foreign
assets can be found in Fedderke and Romm (2006). This analytical framework has the
advantage of being able to handle both steady state, and the dynamics of adjustment to
steady state. Our real concern here is the mix of the two capital assets in the portfolio of
agents.
MODEL SPECIFICATION
Following the analytical framework presented in the previous section, we seek to
econometrically estimate two classes of relationship: (i) the relationship between FDI
inflows and economic growth in Nigeria and (ii) the determinants of FDI inflows to
Nigeria. To ensure that the conclusions we arrive at are robust and useful for policy
making, we employ two alternative estimation techniques for estimating the nature of the
first relationship. Hence, we employ single-equation models and simultaneous equation
models to examine the FDI-growth relationship in Nigeria. Our procedure is motivated
by the notion that there might be simultaneity (bi-directional) bias between FDI and
growth. That is FDI is affected by growth, and at the same time, the rate of economic
growth is influenced by FDI. In the case of the second relationship, (FDI determinants),
we stick to the single equation estimation.
Vol. 11, No.2 Journal of Monetary and Economic Integration
22
Single Equation Models
FDI and Growth
We estimate both the growth impact of FDI and the determinants of FDI by means of a
vector error correction model (VECM). Our model is adapted from Anyanwale (2007);
Akinlo (2004); Ekpo (1995) and Fedderke and Romm (2006). Hence, for the FDI growth
relation, we specify thus: =,,,,,,,,,. . . . . . (11)
In econometric form, the model can be written thus:
=0 +1 + 2 + 3 + 4 + 5 + 6 + 7 + 8 + 9
+ 10+ (12)
Where YG is real GDP growth rate; L is labour, and are stock of private
and foreign capital respectively; Cg is real government consumption, O is trade openness,
H is human capital, D is the adjustment dummy, 1 for adjustment periods 1986-2001 and
0 otherwise, Fn stands for financial depth, Bg is budget balance to GDP and T, is the time
trend to capture the cyclical or secular trends in output during the period under review.
Reparamertrization, and taking lower case letters to denote natural logarithms
and , to denote the difference operator provides the VECM specification:
=0 +1 + 2 + 3 + 4 + 5 + 6
+ 7 + 8 + 9 + 1+ (13)
Where 1 , 2,3 , ... 9 are interpreted as the various elasticises, and 1 is
the short-run error correction coefficient.
FDI Determinants
Here, we specify a model based on the intertemporal optimizing portfolio
analytical framework as developed by Fedderke (2002) and Fedderke and Romm (2006)
and implemented by Ayanwale (2007). Hence, the determinants of FDI inflows are based
on proxies for rates of return and risk in the host country. Thus:
=,,,,,,, (14)
In econometric form,
= 0+ 1+ 2+ 3+ 4+ 5+ 6+ 7+ (15)
Where F is FDI inflows, O is level of trade openness, I is the level of
infrastructure, R is return on investment, N is inflation, G is government size, YP is
GDP per capita, H is human capital and E is real effective exchange rate.
As usual, the process of reparamertrization, and taking lower case letters to
denote natural logarithms and, to denote the difference operator provides the VECM
specification thus:
= 0+ 1 + 2 + 3 + 4 + 5 + 6
+ 7 +1+ 16
Where 0, 1,2,3, ... 7 are the respective elasticises of the variables.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
23
Simultaneous Equation Systems
We first use a five simultaneous equation system and then we narrow down to a two-
equation system to examine the relationship between FDI and growth in Nigeria. The
use of the simultaneous equation estimation is motivated by the simultaneity bias
between FDI and growth.
The three stage least squares (3SLS) approach
We use the three stage least squares approach (Gujarati, 2009) to estimate a system of
five endogenous equations. This approach is appropriate when estimating systems of
equations that are over identified (see Ruxanda and Muraru, 2010; Green, 2003) and it
has been the preferred choice in empirical studies with numerous systems of equations
(see Ghatak and Halicioglu, 2006). The equations of the system are described as follows:
(1) The Growth Equation
This is specified thus:
=1+ 2+ 3+ 4+ 5(17)
Where M stands for the ratio of gross capital formation to GDP, and the other
nomenclatures are as earlier described.
(2) The FDI Equation
This is specified thus:
=6+ 7+ 8+ 9+ 10(18)
(3) Capital Stock Equation
This is specified thus:
=11+ 12+ 13+ 14+ 15(19)
Where is the level of domestic savings and other variables are as described earlier.
(4) Openness Equation
This is specified thus:
=16+ 17+ 18+ 19+ 20(20)
Where E represents the real effective exchange rate, and I is the average deposit rate.
(5) The Savings Equation
This is specified thus:
=21+ 22+ 23+ 24+ 25(21)
All the variables are as earlier described. For each of the equations, we use the lagged
first difference of all the exogenous variables as our instruments. Ruxanda and Muraru
(2010) use a similar approach.
The two stage least squares (2SLS) procedure
In other not to lose a lot of information and degrees of freedom (as it is common with
the 3SLS procedure), we estimate the relationship using a different procedure: i.e. the
Vol. 11, No.2 Journal of Monetary and Economic Integration
24
two stage least squares (2SLS). Thus, we estimate equations (22) and (23) using the 2SLS
procedure. Where,
=1+ 2+ 3+ 4+ 5(22)
=6+ 7+ 8+ 9+ 10(23)
All the variables are as defined earlier. Here again, following Ruxanda and Muraru (2010),
we use the lag values of all the exogenous variables as our instruments.
DATA AND A PRIORI EXPECTATIONS
All the data used for the study was retrieved from four major sources: CBN statistical
bulletin, World Bank World Development Report, IMF International Financial Statistics
(IMF-IFS), UNCTAD Database. Details about the sources, description and construction
of the datasets can be found in the Appendix section. To determine the a priori
expectations, we observe the signs and magnitudes of the parameters to be estimated.
We place emphasis on Eq. (13) and Eq. (3.15). The expected signs and magnitudes in
these two equations are also valid for all the parameters in the simultaneous equation
systems. In the FDI-growth equation (i.e. Eq. 13), we expect the signs of the parameters
1 , 2, 3, 4 and 7 to assume positive signs. This theoretical expectation follows
naturally from the analysis of production theory. 1 represents the coefficient for labour,
and the higher the labour input in a production process, the higher will be the output.
2 represents the stock of private capital, while 3 is for foreign capital. Again, from the
simple production function, the higher the capital input in a production process, the
higher will be the level of output, hence, our theoretical expectation of a positive sign for
these parameters are justified.
We also anticipate that the parameter 4 will assume a positive sign. Thus is in line with
theories of human capital development which postulates that the better the quality and
supply of human capital the higher will be the productivity of labour. The parameter 5
is indeterminate. This is because there is no straight rule about the effects of openness
on an economy. Openness may harm or accelerate growth in an economy depending on
the level of development of that economy. The effect of government consumption on an
economy is not also certain. It depends on whether or not government expenditure
crowds-out private consumption. If government expenditure crowds-out private
consumption, then 6 will be positive. If it does not, then it will be negative. The sign of
the parameter for financial development ( 8) is indeterminate. It depends on whether
financial development reduces or increases capital flight. If it increases capital flight, it
will have a positive value. If it does not, it will have a negative value (Akinlo, 2006). The
coefficient for the adjustment variable, ( 9) may take a positive or negative sign,
depending on the way the adjustment works. If adjustment enhances efficiency, as it
should, the sign should be positive. A negative sign for the adjustment variable will imply
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
25
that the SAP programme was not growth enhancing. ( 10) can be positive or negative,
depending on whether annual growth rate in the country increased or decreased during
the period.
RESULTS AND DISCUSSION
INTEGRATION AND COINTEGRATION PROPERTIES OF THE FDI-
GROWTH EQUATION
Table 3 presents the unit root tests for all the variables used in the equations. The test is
conducted using two different unit root models. That is, the Augmented Dickey Fuller
(ADF) model and the Philips-Perron (PP) model. The essence of using the two testing
procedures is for confirmatory testing.
Table 3: Unit Root Tests
Variables
ADF PP
Level 1st Difference Level 1st Difference Conclusion
BG -3.91(0) *** - -3.97(0)*** - I(0)
E -0.99(1) -5.73(0)*** 0.99(0)* -5.34(1)*** I(1)
F 4.98(0)*** - 4.94(1)*** - I(0)
FN -0.85(1) -5.14(0)*** 0.91(5) 6.08(3)*** I(1)
I -0.67(0) -7.76(0)*** 0.69(5)*** 8.10(4) *** I(1)
H -2.892(0)* - 0.82(5) 3.24(2) *** I(1)
L 0.33(0) -4.32(0)*** 0.81(5) 4.26(1) *** I(1)
KP 0.89(1) 5.45(0)*** 0.75(0) 6.04(0)*** I(1)
N -1.85(0) 11.45(2)*** 1.72(1) -8.31(1)*** I(1)
O 1.52(2) 8.21(1)*** 1.02(2) -9.01(1)*** I(1)
R -1.93(9) -6.63(8)*** 1.82(5) 5.23(2) *** I(1)
Y 0.21(1) 7.52(0)*** 0.81(5) -7.28(1) *** I(1)
YG -1.02(0) 4.23(0)*** -1.75(0) -3.04(0)*** I(1)
YP -0.25(2) -3.01(0)*** 1.02(1) 8.31(1)*** I(1)
Notes: ***,** and * indicates significance at the 1%, 5%, 10% levels respectively. The values in
bracket for the ADF and PP test, indicates the optimal lag length selected by the SIC within a
maximum lag of 13.
Vol. 11, No.2 Journal of Monetary and Economic Integration
26
The Schwarz Information Criterion (SIC) is used to select the optimal lag length of the
models. The tests are conducted with a maximum permissible lag length of 9 lags. Table
3 indicates that all the variables in the model are not stationary at the levels except for
two variables: fiscal budgets as a ratio of GDP and foreign direct investment. After
taking the first differences of all the other variables they became stationary. Interestingly,
the two tests statistics (ADF) and (PP) returned results that lead to similar conclusions.
These results imply that the regression results that would be obtained from the models
specified in Chapter 3 would return spurious results if there is no long-run relationship
among the variables in the model. Since not all the variables are stationary at levels, it
necessarily means that we have to investigate the cointegration properties of the variables
in the equations. The results for the cointegration tests are presented in Table 4.
Table 4: Johansen Cointegration test for FDI-Growth Model
Null Hypothesis Alt. Hypothesis Test Statistic Critical Value (5%) P-Value
Trace test
r = 0 r = < 1 549.6560 197.3709 0.0001
r = 1 r = < 2 367.5954 159.5297 0.0000
r = 2 r = < 3 248.4119 125.6154 0.0000
r = 3 r = < 4 163.6921 95.75366 0.0000
r = 4 r = < 5 96.17374 69.81889 0.0001
r = 5 r = < 6 54.21201 47.85613 0.0113
r = 6 r = < 7 31.41903 29.79707 0.0322
r = 7 r = < 8 17.94208 15.49471 0.0210
r = 8 r = < 9 4.709390 3.841466 0.0300
Maximum Eigenvalue test
r = 0 r = 1 182.0606 58.43354 0.0000
r = 1 r = 2 119.1835 52.36261 0.0000
r = 2 r = 3 84.71986 46.23142 0.0000
r = 3 r = 4 67.51833 40.07757 0.0000
r = 4 r = 5 41.96173 33.87687 0.0044
r = 5 r = 6 22.79297 27.58434 0.1825
r = 6 r = 7 13.47696 21.13162 0.4093
r = 7 r = 8 13.23269 14.26460 0.0723
r = 8 r = 9 4.709390 3.841466 0.0300
Trace test indicates 9 cointegration equation while the maximum eigen value test indicates 5
cointegrating equations at the 5% level of significance.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
27
We utilize the Johansen co-integration test procedure and use both the Trace criterion
and the Maximum Eigenvalue criterion to determine the rank of the cointegrating
relationships among the variables. The decision criterion is thus: when the Trace Statistic
is greater than the 5% critical value, we reject the null hypothesis of no cointegrating
relation and conclude that there is cointegration among the variables. We continue the
testing in an iterative manner until we are no longer able to reject the null hypotheses of
no cointegrating relationship. Table 4 presents the Unrestricted Cointegration Rank Test
using the Trace Statistic. The test is conducted with the assumption that there is a trend
and a constant term in the model. Further, the SIC criterion is used to determine the
optimal lag length of the cointegrating VAR equations.
The trace test indicates that there are 9 cointegrating equations in the FDI-growth
equation, while the maximum eigenvalue test indicates that there are 5 cointegrating
equations in the model. These results lead to the conclusion that there is a long- run
relationship among the variables in the equation. It is economic commonsense that long-
run relationships usually have disequilibrium in the short-run; hence, to tie the short-run
distortions in the relationship to the long-run equilibrium relationship, it is necessary to
estimate an error-correction model, which will show the speed of adjustment and the
average time it will take for short-run distortions in the relationship to be corrected
(Chuku, 2009). The results of these models are presented in the next section.
ANALYSIS OF THE RESULTS FROM THE SINGLE – EQUATION FDI-
GROWTH MODEL
Table 5 presents the results of the single equation FDI-Growth model. The results are
presented with the error correction specification side-by-side. From the table, it can be
observed that some of the apriori expectations do not hold.
Vol. 11, No.2 Journal of Monetary and Economic Integration
28
Table 5: Estimates of the Standard and ECM FDI-Growth Equation
Dependant Variable GDP growth rate (YG)
Standard Model ECM Model
Constant -1.660088 -0.502262
(3.68013) (3.5542)
t 0.1127*** 5.78E-06**
(0.2148) (1.81E-05)
D(L) 5.53E-07** 6.00E-06**
(6.53E-05) (2.93E-05)
D(KP) -1.32E-06** -2.30E-07
(-3.28E-05) (1.74E-05)
D(F) 5.9923*** 5.49029***
(1.91432) (1.49326)
D(H) 1.53E-06 0.025996
(1.74E-05) (0.0107)
D(O) -0.0078*** -0.539213
(0.002556) (0.0411)
D(CG) -0.030715 -0.04879
(0.046470) (0.1064)
D(BG) 0.485657 -1.2490
(0.317018) (3.02124)
D(FN) 0.011264** 0.07518**
(0.02574) (0.02040)
SAP -2.029251 0.075187
(3.396944) (0.0107)
ECM(-1)
-0.5168***
(0.1843)
R2 Adj. 0.1198 0.3125
D.W 2.89 2.61
ARCH (2 lags) 0.42 (0.6557) 13.41 (0.000)
RESET (2 terms) 0.61 (0.5517) 7.62 (0.0018)
Chow (1986) 2.03 (0.09173) 2.03(0.0913)
Notes: ***, ** and * denotes asymptotic significance at the 1%, 5%, and 10% levels respectively.
Values in brackets represent standard errors for parameters and p-values for the relevant test statistic.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
29
Column two of Table 5 presents the results from the standard single equation FDI-
Growth-model. Model specification tests for the standard model are presented in the
lower rows of the column. The adjusted R squared value of 11.98% is an indication that
the model is a poor-fit of the relationship between economic growth and foreign direct
investment in Nigeria. The Durbin Watson statistic of 2.89, which is well above 2, is also
suggestive that there is positive first-order auto correlation in the error terms from the
equation. The ARCH test indicates that there is autoregressive conditional
heteroscedasticity in the error terms. Finally, the regression specification (RESET) test
for omitted variables cannot be rejected; leading to the conclusion that the model may
have omitted some relevant variables. For these reasons and others, it is irrelevant to
discuss the results of the FDI-Growth standard model. The results from the error
correction model are rather discussed.
Column 3 presents the results of the error correction specification of the FDI-Growth
model. The second coefficient t represents the impact of technological development on
FDI. A time trend was used as a crude proxy for technological development. The
positive and significant sign of the coefficient is an indication that technological
developments have had a positive impact on the level of economic growth in Nigeria.
This can also be interpreted as evidence of technology diffusion which can be attributed
to foreign direct investment.
As expected, labour force has a positive relationship with economic growth in Nigeria.
The parameter estimate for labour force is significant at the 5% level. This result
conforms with traditional growth models which speculate that increases in material input
in the production process leads to increases in material output. Further, private capital
(Kp) assumed a positive and insignificant value. The most crucial parameter in the
model, that is, the parameter for foreign direct investment assumed a positive and
significant value. This result is in line with the a priori expectations of the ownership,
location and internalization (OLI) theory of foreign direct investment. This result
provides strong evidence of the positive and significant impacts of FDI on economic
growth in Nigeria.
Though not statistically significant, the parameter estimates for trade openness indicates
that trade openness has been growth inhibiting in Nigeria. This result can pose policy
challenges and it is also difficult to reconcile the fact that FDI causes growth, whereas,
openness inhibits growth. Though this kind of inconsistency in results can be attributed
to data issues, it is however intuitive for the policy maker, because the results may imply
that the nature of FDI-inflows to Nigeria may not have been evenly distributed in such a
way that forward and backward linkages can be generated. Indeed Anyanwale (2007)
found in his study on economic growth and FDI-inflows to Nigeria, that most of the
FDI that came to Nigeria was directed to the extractive industries. This is an indication
of the weak interactions that exist between FDI policies and trade policies in Nigeria.
Vol. 11, No.2 Journal of Monetary and Economic Integration
30
Government size as a ratio of GDP (Cg) and the budget balance as a ratio (Bg) of GDP,
both assumed negative signs. This implies that fiscal activities of government in Nigeria
have had negative impacts on growth in Nigeria. This outcome may not be surprising
because, the domination of government‟s fiscal activities (consumption and production)
in the economy is usually not efficiently done, thereby not producing the required effects
on growth. Another implication of these negative signs is that government consumption
and fiscal activities have some kind of crowding out effect on private consumption and
investment behaviours. All of these may explain the reason for the inverse relationship
observed between government consumption as a ratio of GDP and government fiscal
balance as a ratio of GDP, and economic growth in Nigeria. Interestingly, the parameter
estimate for the effect of financial structure (FN) on economic growth assumed a
positive sign as expected. Though the estimate is not statistically significant, it gives us a
vague picture of the positive impacts of financial development on economic growth in
Nigeria.
The parameter estimate of utmost interest is that of the error correction variable, which
seeks to tie short run distortion in the model to the long-run equilibrium path. The ECM
variable used in the model, are the residual terms derived from the standard FDI-growth
model in Column 2, Table 5. The residuals, as is conventional in econometric practice,
are introduced into the ECM model after taking the first lag. From the table, we observe
that the parameter estimate is -0.5168, and the estimate is significant at the 1% level of
significance. This value can be interpreted to mean that 51.68% of the distortions in the
equilibrium relationship between economic growth and the variables included in the
model would be corrected in the present period. To find out how long it will take to fully
correct any distortion in the long-run relationship, we simply divide one by the ECM
coefficient i.e. (1/0.5168) = 1.934. Since this value is close to two, it implies that it will
take approximately two years for full adjustments to take place after a shock has
occurred.
A quick examination of the model specification tests presented in the lower rows of the
Table 5 clearly indicates that the ECM model is superior to the standard model. First, the
adjusted R squared value of 13.41, though still poor, better fits the data than the standard
model. The Durbin-Watson statistic of 2.6 is a weak indication of the presence of
positive serial autocorrelation in the residuals (which can be ignored). The ARCH test
clearly leads to the rejection of the null hypothesis of the presence of auto regressive
conditional heteroscedasticity in the models. The RESET test also suggests that the
variable no longer suffers from omitted variable bias.
Tests of structural breaks within the period under review are also carried out. From the
results obtained in the standard and ECM models, it is clear that the Structural
Adjustment Programme (SAP) which commenced in 1986, caused some kind of
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
31
structural break in the relationship between economic growth and the variables included
in the model. This can be observed from the F values of 2.03 and 2.04 of the Chow
breakpoint test with statistically significant probabilities.
ANALYSIS OF THE RESULTS FROM THE FDI DETERMINANTS
EQUATION
From Table 3, which presents the unit root properties of the variables, we observe that
the variables that belong to the FDI determinants model are also I(1) stationary, i.e. they
become stationary after we took the first differences of the series. This also implies that
we need to investigate the cointegration properties of the models before we proceed with
the model estimation. Table 6 summarizes the results of the cointegrating relationships
among the variables. The trace test indicates that there are 8 cointegrating equations in
the model while the maximum Eigen-value test indicates that there are 4 cointegrating
equations. These results imply that we can go ahead with the estimation of the model
without any fear of obtaining spurious regression results. It also implies that the error
correction mechanism (ECM) can be used to tie the short-run disequilibrium in the
relationship with the long-run values. Table 7 presents the results from the FDI
determinants model. Again, the standard and ECM specifications of the model are
presented side-by-side.
Table 6: Johansen Cointegration test for FDI Determinants Model
Null Hypothesis Alt. Hypothesis Test Statistic
Critical Value
(5%) P-Value
Trace test
r = 0 r = < 1 400.1041 197.3709 0.0000
r = 1 r = < 2 270.3922 159.5297 0.0000
r = 2 r = < 3 184.9574 125.6154 0.0000
r = 3 r = < 4 137.0141 95.75366 0.0000
r = 4 r = < 5 96.47311 69.81889 0.0001
r = 5 r = < 6 62.91786 47.85613 0.0011
r = 6 r = < 7 32.61723 29.79707 0.0231
r = 7 r = < 8 15.78053 15.49471 0.0453
r = 8 r = < 9 3.458285 3.841466 0.0629
Vol. 11, No.2 Journal of Monetary and Economic Integration
32
Maximum Eigenvalue test
r = 0 r = 1 129.7119 58.43354 0.0000
r = 1 r = 2 85.43488 52.36261 0.0000
r = 2 r = 3 47.94326 46.23142 0.0325
r = 3 r = 4 40.54099 40.07757 0.0443
r = 4 r = 5 33.55524 33.87687 0.0546
r = 5 r = 6 30.30063 27.58434 0.0218
r = 6 r = 7 16.83670 21.13162 0.1798
r = 7 r = 8 12.32225 14.26460 0.0991
r = 8 r = 9 3.458285 3.841466 0.0629
Trace test indicates 8 cointegration equation while the maximum eigenvalue test indicates 4 cointegrating
equations at the 5% level of significance
Here, the results from the ECM model are analyzed with references made to the results
from the standard model where necessary. The result shows that openness has a positive
effect on the inward flows of foreign direct investments in Nigeria. This outcome is as
expected. Specifically, the parameter value of 0.0127 implies that a one unit increase in
the level of openness of the economy will on average bring about 1.27 units of increase
in FDI inflows to Nigeria. The explanation for this relationship is simple, because the
more open an economy is, the easier it will be for capital and financial commodities to be
exchanged across boundaries.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
33
Table 7: Estimates of the FDI Determinants Equation
Dependant Variable: First Difference of the log of FDI, D(LOG(FDI))
Standard Model ECM Model
Constant 0.1988 0.1791
(0.1480) (0.1564)
D(O) 0.0943*** 0.0127**
(0.0058) (0.0037)
D(I) -0.0165** -0.1416
(0.0071) (0.3929)
D(R) -0.5113*** -0.04298***
(0.0058) (0.0019)
D(N) 0.0255** 0.0295
(0.0040) (0.0411)
D(C
g
) 0.0625 0.0861**
(0.0023) (0.0089)
D(YP) 5.82E-08 0.0876
(0.0018) (1.26E-07)
D(H) 0.7137*** 5.55E-08
(1.1921) (0.013273)
D(E) -0.0065 0.1595**
(0.0121) (0.0137)
ECM(-1)
-0.2192***
(0.0615)
R2 Adj. 0.059 0.145
D.W 2.46 1.91
ARCH (2 lags) 1.35 (0.2751) 1.11 (0.3428)
RESET (2 terms) 0.71 (0.6517) 7.62 (0.0018)
Chow (1986) 0.873 (0.56099) 0.74(0.6727)
Notes: ***, ** and * denotes asymptotic significance at the 1%, 5%, and 10% levels respectively.
Values in brackets represent standard errors for parameters and p-values for the relevant test statistic.
Vol. 11, No.2 Journal of Monetary and Economic Integration
34
A second variable that significantly influences the inflows of foreign direct investments
to Nigeria is the global rate of return on investments (R). The parameter estimate for R
assumed a negative and significant value at the 1% level of significance. The negative
sign is an indication that as the rate of return on investments increases in the rest of the
world, FDI inflows in Nigeria will reduce. This is behaviourally plausible since investors
are likely to take their funds to highest yielding locations. Government consumption as a
ratio of GDP (Cg) is also a significant determinant of FDI inflows to Nigeria. Since,
government is the dominant player on the macroeconomic field in Nigeria; investors may
be motivated or demotivated by government fiscal activities in arriving at decisions to
invest in the economy. Specifically, the result shows that a one unit increase in the ratio
of government consumption to GDP leads on average to an 8.61% increase in the inflow
of foreign direct investment. Surprisingly, our results show that the quality of human
capital available in Nigeria has also been a significant determinant of FDI inflows to
Nigeria. This is because the coefficient for variable E assumed a positive and significant
sign. The rationale for this relationship is thus: investors will be more and more willing to
bring in their capital to invest in Nigeria if they can be sure that the labour force is
developed enough to easily internalize and diffuse the new and modern techniques of
production that would be introduced, thereby, removing the need for importing human
capital from the originating firms‟ country. As expected, the real effective exchange rate
is a significant determinant of FDI inflows to Nigeria. Since the other variables included
in the model are not statistically significant, we conclude that they are not major
determinants of FDI inflows into Nigeria.
As expected, the coefficient of the error correction model assumed a negative and
statistically significant sign. The ECM coefficient informs us that 21.92% of the
distortions in the long-run relationship among the variables are corrected in the current
period, and it takes approximately 4.56 years to attain full adjustment back to the
equilibrium relationship. The adjusted R squared value of 0.145 indicates that the model
does not satisfactorily fit the data set. Other model specification tests, such as the D.W
statistic indicates that there might be negative serial autocorrelation in the residuals of the
estimated regression line. Further, the ARCH test is not satisfactory, whereas, the
RESET test indicates that the variable does not suffer from the common omitted
variable bias usually observed in models with low R squared values (Wooldridge, 2006).
The Chow breakpoint test does not provide any evidence that could suggest that there
was any structural break caused by the SAP in 1986.
To ascertain whether the results obtained in the model can be of any use for policy
analysis, we carry out coefficient stability test of the standard and ECM model using the
cumulative sum (CUSUM) test and the cumulative sum of squares test (CUSUM
Squared). Figure 1 presents the charted result of the CUSUM test, while Figure 2
presents the result of the CUSUM of Squares test. To read the chart, one simply needs to
observe the movement of the cumulative sum of squares, if it lies within the 95%
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
35
confidence interval, then we conclude that the parameters are stable, and hence the
conclusion that emerge from the estimation procedure can be useful for policy analysis.
Here, we observe that using the two competing tests (i.e. CUSUM and CUSUM of
Squares test), the sum of squares lies within the bands of 95% interval, indicating that the
coefficients are stable.
Figure 1 CUSUM Test
Figure 2 CUSUM of Squares Test
-15
-10
-5
0
5
10
15
84 86 88 90 92 94 96 98 00 02 04 06 08
CUSUM 5% Signi ficance
-0.4
0.0
0.4
0.8
1.2
1.6
84 86 88 90 92 94 96 98 00 02 04 06 08
CUSUM of Squares 5% Significance
Vol. 11, No.2 Journal of Monetary and Economic Integration
36
ANALYSIS OF SIMULTANEOUS EQUATION MODELS
THREE STAGE LEAST SQUARES (3SLS) ESTIMATES
The method of three stage least squares estimation is popular in applied work because it
is known to be fully efficient since it takes into account all available information in the
estimation of the coefficients of a model, and then forms weights and re-estimates all
the coefficients of the model using the estimated weighting matrix. The use of this
approach is appropriate when the right hand side variables are correlated with the error
terms and there is both heteroscedasticity and contemporaneous correlation in the
residuals (Gujarati and Porter, 2009; Eviews, 5.1 Userguide). The summary of the results
obtained by using the 3SLS method are presented in Table 8. In Table 9, the details of
the five equations with their instrumental variables are presented. The coefficient
specification and the respective results for the second-order tests, i.e. the econometric
criteria such as the R2 value, D.W and other criteria are also displayed in the table.
The coefficients C(1) to C(5) represents the parameter estimates obtained from the
growth equation. C(6) to C(10) are the coefficients obtained from the FDI equation,
C(11) to C(15) represents the estimates for the variables in capital stock equation. C(16)
to C(20) are for the openness equation and C(20) to C(25) are for the savings function.
From the growth equation, it can be observed that only three of the coefficients are
statistically significant, i.e. C(1), C(2) and C(4). Since C(1) is the constant term, we
concentrate on explaining the implications of C(2) and C(4) being significant. C(2)
represents the parameter estimate for FDI, the positive sign it assumes implies that there
is a direct relationship between FDI inflows to Nigeria and economic growth. The
relationship is such that a one unit increase in the inflow of FDI to Nigeria will
approximately lead to a 0.02 units increase in the level of growth. Also, the coefficient
C(4) shows that there is a negative relationship between the level of openness in the
economy and economic growth in Nigeria. The relationship is such that a one unit
increase in the degree of openness of the economy will on average lead to 3.62 units of
decrease in the growth rate.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
37
Table 8: Estimates from three-stage least squares regression
Endogenous Variables: (YG), F, M, O, SD
Coefficient Std. Error t-Statistic Prob.
C(1) 31.12033 15.23691 2.023372 0.0469
C(2) 0.020170 0.011500 2.340318 0.0341
C(3) 0.004654 0.026319 0.176847 0.8599
C(4) -3.620234 1.853707 -3.634591 0.0084
C(5) -1.51E-05 0.000104 -0.145354 0.8846
C(6) -243461.0 411882.5 -0.591093 0.5553
C(7) 1732940. 2995535. 4.578508 0.0008
C(8) -60.76414 284.5733 -0.213527 0.8312
C(9) 3233.385 2031.695 4.591472 0.0054
C(10) -0.786862 1.021263 -0.770480 0.4422
C(11) 5.401337 642.9134 0.008401 0.9933
C(12) 3.50E-05 0.009056 3.003869 0.0009
C(13) -46.71519 1718.352 -0.027186 0.9783
C(14) 7.327896 3.132395 2.338379 0.0003
C(15) 6.884509 4. 378795 -1.913439 0.0893
C(16) 287.4563 385.9026 0.744893 0.4575
C(17) 1159.835 2492.008 1.465422 0.0423
C(18) -17.73678 33.18012 -0.534560 0.5937
C(19) -0.167051 4.646458 -0.035952 0.9714
C(20) 1.505825 10.30836 0.146078 0.8841
C(21) 10.96158 446.2230 0.024565 0.9804
C(22) 68.60937 27. 45336 -2.024991 0.0001
C(23) 0.171471 7.451989 0.023010 0.9817
C(24) -0.404457 20.42054 -0.019806 0.9842
C(25) -5.57E-06 0.000294 -0.018937 0.9849
Determinant residual variance 2.40E+28
Author’s computations
Vol. 11, No.2 Journal of Monetary and Economic Integration
38
Looking at the FDI equation, we observe that the parameter estimates for the GDP and
openness variable (i.e. C (7) and C (9)) are the only statistically significant variables. C(7)
tells us that there is a positive relationship between the level of output and the inflows of
FDI in Nigeria. This kind of relationship follows the speculations of the market-seeking
theory of FDI (see Chapter 2). Similar to the relationship obtained in the single equation
model, we also observe that the degree of trade openness has a positive relationship with
the inflows of FDI into Nigeria.
The capital stock equation, though not of primary concern in this work still contains
interesting results that is worth examining. From the capital stock equation, we observe
that three variables significantly affect the level of capital stock in Nigeria. They include,
foreign direct investment, the financial system and the savings to GDP ratio represented
by C(12), C(14) and C(15) respectively. As expected, foreign direct investment has a
significant and positive influence on the capital stock in Nigeria. This is theoretically
plausible as foreign direct investment is expected to bridge some of the savings gap that
may not be obtained domestically. Also, developments in the financial system including
financial deepening positively affect the level of capital stock in the economy. The
relationship is such that a 1 unit increase in the level of financial deepening in the
economy will lead to approximately 7.32 units of increase the level of capital stock in the
economy. Further, the rate of savings in the economy directly affects the level of capital
stock in the economy. This follows the expectation of the Keynesian models of savings
and investments.
Surprisingly, the openness equation tells the story that only the level of output (i.e. GDP)
in the economy significantly influences the level of openness in Nigeria. This conclusion
is derived from the estimated value of C(17) which is 1159.83. Finally, the fifth equation,
which represents the savings equation also shows that the only significant variable that
influences savings in Nigeria is the level of output. This again conforms to the simple
Keynesian models of National income, where it is assumed that savings is an increasing
function of disposable income.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
39
Table 9: 3SLS specification and summary measures
Growth Equation: D(YG) = C(1) + C(2)*D(F) + C(3)*D(M) + C(4)*D(O) +C(5)
*D(H)
Instruments: D(F(-1)) D(M(-1)) D(O(-1)) D(YG(-1)) C
Observations: 36
R-squared -140.802979 Mean dependent var -0.216325
Adjusted R-squared -159.100128 S.D. dependent var 7.535132
S.E. of regression 95.34254 Sum squared resid 281796.2
Durbin-Watson stat 1.979974
FDI Equation: D(F) = C(6) + C(7)*D(LOG(Y)) + C(8)*D(M) + C(9)*D(O)
+C(10)*D(H)
Instruments: D(LOG(Y(-1))) D(M(-1)) D(O(-1)) D(H(-1)) F(-1) C
Observations: 35
R-squared -9.808422 Mean dependent var 63696.09
Adjusted R-squared -11.249545 S.D. dependent var 232059.7
S.E. of regression 812193.9 Sum squared resid 1.98E+13
Durbin-Watson stat 2.066490
Capital Stock Equation: D(M) = C(11) + C(12)*D(F) + C(13)*D(LOG(Y)) + C(14)*D(FN)
+ C(15)*D(SD)
Instruments: D(F(-1)) D(LOG(Y(-1))) D(FN(-1)) D(SD(-1)) C
Observations: 36
R-squared -0.003299 Mean dependent var 2.741945
Adjusted R-squared -0.132757 S.D. dependent var 961.1520
S.E. of regression 1022.964 Sum squared resid 32440118
Durbin-Watson stat 2.999569
Openness Equation: D(O) = C(16) + C(17)*D(LOG(Y)) + C(18)*D(E) + C(19)*D(M)
+ C(20)*D(I)
Instruments: D(F(-1)) D(LOG(Y(-1))) D(E(-1)) D(I(-1)) C
Observations: 35
R-squared -10.347033 Mean dependent var 64.56400
Adjusted R-squared -11.859970 S.D. dependent var 153.6865
S.E. of regression 551.1321 Sum squared resid 9112397.
Durbin-Watson stat 2.233819
Savings Equation: D(SD) = C(21) + C(22)*D(LOG(Y)) + C(23)*D(I) + C(24)*D(E)
+C(25)*D(H)
Instruments: D(SD(-1)) D(LOG(Y(-1))) D(E(-1)) D(I(-1)) C
Observations: 36
R-squared -89.309799 Mean dependent var 0.239859
Adjusted R-squared -100.962677 S.D. dependent var 2.782814
S.E. of regression 28.09990 Sum squared resid 24477.73
Durbin-Watson stat 1.927645
Author’s computations
Vol. 11, No.2 Journal of Monetary and Economic Integration
40
A quick examination of the econometric criterion of the model reveals that there is
absence of serial auto correlation in the residuals of the estimate. This is because the
D.W statistic for the five equations as presented in Table 9 ranges from 1.92 to 2.99
which are both close to 2. We neglect the negative R squared values obtained for the
various equations. This is in line with standard econometric practice which is clearly
stated by Wooldridge (2006), thus: “if we use 3SLS residuals to compute the SSR for
both the restricted and unrestricted models, there is no guarantee that SSRr> SSRur; if the
reverse is true, then the F statistic and the R squared statistic will be negative, thereby
rendering inappropriate, the usual way of interpreting these statistics” Overall, the
results from the single equation estimation of the FDI-growth model and the results
obtained from the 3SLS estimation procedure do not vary significantly. This may be a
crude indication of the absence of simultaneity bias in the nature of the relationship
between economic growth and foreign direct investments in Nigeria.
Two Stage Least Squares Estimates
Using the three-stage least square approach have been described as an over
parameterized model because it uses a lot of information that may not be useful to the
structural equations that are being focused on. Also, in other not to lose a lot of
information and degrees of freedom (as it is common with the 3SLS procedure), we
estimate the relationship using a different procedure: i.e. the two stage least squares
(2SLS) approach. This approach allows us to focus on the two relationships of key
interest, (i.e., growth and FDI as endogenous variables respectively) and it is also
superior to the single equation ordinary least squares method because the estimates from
2SLS are unbiased and consistent (Gujarati and Porter, 2009).
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
41
Table 10: Estimates from two-stage least squares regression
Endogenous Variables: YG, F
Coefficient Std. Error t-Statistic Prob.
C(1) -228603.4 247395.7 -0.924040 0.3590
C(2) 358189.7 135129.7 2.265071 0.0301
C(3) 3865.340 1915.599 2.017823 0.0479
C(4) 17.49634 233.9992 0.074771 0.9406
C(5) -0.017869 0.811158 -0.022029 0.9825
C(6) 1.565685 3.266198 0.479360 0.6334
C(7) 1.35E-05 3.95E-05 2.342533 0.0331
C(8) 0.643428 0.616532 1.043625 0.3007
C(9) -2.69E-06 4.27E-06 -0.630673 0.5306
C(10) -0.021546 0.181408 -0.118770 0.9058
Determinant residual
covariance 2.21E+13
Equation: D(F) = C(1) + C(2)*D(LOG(Y(-1))) + C(3)*D(O) + C(4)*D(M)
+C(5)*D(H)
Instruments: D(F(-1)) D(M(-1)) D(O(-1)) D(YG(-1)) C
Observations: 36
R-squared -4.855107 Mean dependent var 61931.05
Adjusted R-squared -5.610605 S.D. dependent var 228965.6
S.E. of regression 588695.6 Sum squared resid 1.07E+13
Durbin-Watson stat 2.017029
Equation: D(YG) = C(6) + C(7)*D(F) + C(8)*D(BG) + C(9)*D(H) + C(10)
*D(FN)
Instruments: D(M(-1)) D(O(-1)) D(H(-1)) D(F(-1)) D(LOG(Y(-1))) C
Observations: 36
R-squared -0.427402 Mean dependent var 0.053360
Adjusted R-squared -0.611583 S.D. dependent var 7.313291
S.E. of regression 9.284088 Sum squared resid 2672.023
Durbin-Watson stat 2.706788
Author’s computations
Vol. 11, No.2 Journal of Monetary and Economic Integration
42
Table 10 contains the parameter estimates obtained by using the 2SLS technique. The
endogenous variables for the model are economic growth and FDI inflows. The
coefficients from C(1) to C(5) are the parameter estimates for the FDI equation while the
coefficients C(6) to C(10) are the parameter estimates for the growth equation. The
result from the FDI equation validates the results that were obtained using the single
equation and 3SLS methods of estimation. That is, that FDI significantly contributes to
economic growth in Nigeria. This result is obvious from the C(2) parameter estimate in
Table 10. The results also show that trade openness is positively related to FDI inflows
in Nigeria; the relationship is such that if the volume of trade is increased by a thousand
units, then FDI will be increased by 3865.34 units. Contrary to the results obtained using
the previous methods of estimation, the 2SLS method returns results that may suggest
that the level of human capital development is negatively related to the level of FDI
inflows in Nigeria. This result is contrary to theoretical expectations, especially following
the internalization theory of FDI. Since the parameter estimate for human capital
development C(5) is not statistically significant, we prefer not to investigate this result
further.
The output from the growth equation is also in conformity with the results using the
single equation and 3SLS approaches. Here, the value of C(7) implies that FDI positively
and significantly affects the rate of economic growth in Nigeria. Since FDI inflows
creates more investments and jobs and reduces the savings-investment gap, it then
follows that it enhances growth. Though the results show that government size
influences growth positively, it however is not statistically significant at the 10% level of
significance. Another coefficient that is worth noting is the parameter estimate for
financial deepening which shows that there is a negative relationship between financial
deepening and economic growth. This result is contrary to theoretical expectations. The
explanation for this relationship can be attributed to data measurement and reliability
problems. We do not wish to pursue this point any further as the C(10) estimate is not
statistically significant.
The conclusions arrived at from the results above are suggestive that there is truly some
form of simultaneity between economic growth and foreign direct investments inflows in
Nigeria. But an intuitive conclusion of the existence of simultaneity may not be sufficient
to prove if there is some kind of bi-directional relationship between economic growth
and foreign direct investments inflows to Nigeria. This motivates us to conduct tests of
endogeneity as specified by the Hausman‟s test. These tests are presented in the next
section.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
43
HAUSMAN’S TEST FOR ENDOGENEITY
The usefulness of the 2SLS estimator is conditional on the endogeneity of the
explanatory variables. Therefore, it is useful to have a test for endogeneity of the
explanatory variable that shows weather the adoption of 2SLS is necessary (Wooldridge,
2006). Hausman (1978) suggested directly comparing the OLS and 2SLS estimates and
determining whether the differences are statistically significant, or by estimating the
reduced form equation of the endogenous variable by regressing it on all the other
exogenous variables (including those in the structural equations and the instrumental
variables), and then obtaining the residuals, which will be added to the structural
equation again, and then the statistical significance of the residuals will be tested using
and OLS regression. If the coefficient of the residual term is statistically significant, we
conclude that the variable is indeed endogenous (Wooldridge, 2006). We follow
Hausman‟s specification in testing whether FDI and economic growth are both
endogenous in the system. Table 11 and Table 12 present the first and second stage
results for the Hausman‟s specification test for endogeneity on foreign direct investment.
Table 11 contains the first stage of the test; it contains the results of the OLS estimation
of foreign direct investment on all other exogenous variables in the model, except
economic growth.
Table 11: First Stage Estimates in the Hausman Endogeneity Test
Dependent Variable: D(F)
Variable Coefficient Std. Error t-Statistic Prob.
C 46284.12 35911.39 1.288842 0.2084
D(L) -0.356040 0.652475 -0.545676 0.5898
D(KP) -0.645883 0.300102 -2.152212 0.0405
D(H) 0.025647 0.027039 0.948498 0.3513
D(O) 21.41647 121.5450 0.176202 0.8615
D(CG) -293.7612 463.7660 -0.633425 0.5318
D(BG) -2183.016 3159.445 -0.690949 0.4955
D(FN) 5609.176 440.7076 12.72766 0.0000
SAP 32567.16 33572.38 0.970058 0.3406
T -5552.035 1876.839 -2.958184 0.0064
Vol. 11, No.2 Journal of Monetary and Economic Integration
44
R-squared 0.902950 Mean dependent var 69442.33
Adjusted R
-squared 0.870600 S.D. dependent var 230340.0
S.E. of regression 82858.48 Akaike info criterion 25.71312
Sum squared resid 1.85E+11 Schwarz criterion 26.14850
Log likelihood -465.6926 F-statistic 27.91182
Durbin-Watson stat 1.647073 Prob(F-statistic) 0.000000
Author’s computations
After estimating the first equation, the residuals from this regression are obtained, and
then the original growth model is re-estimated with the obtained residuals as one of the
explanatory variables. The results from this second stage of regression are presented in
Table 12. The variable of interest in this regression equation is the parameter estimate of
the residual from the FDI equation presented above. This variable is labelled as RES-F.
The parameter estimate of the residual variable is statistically different from zero at the
5% level of significance. This can be easily deduced from its probability value of 0.025.
The statistical significance of the residuals is an indication that FDI and economic
growth are jointly determined, in other words, they are truly endogenous in the equation,
thereby, justifying the estimation of the equations by means of 2SLS regression.
Table 12: Second Stage Estimates in the Hausman Endogeneity Test
Dependent Variable: D(YG)
Variable Coefficient Std. Error t-Statistic Prob.
C 1.223863 6.368191 0.192184 0.8491
D(L) -2.27E-05 7.25E-05 -0.313835 0.7562
D(KP) -5.34E-05 6.96E-05 -0.767270 0.4498
D(F) -6.24E-05 0.000103 -0.608281 0.5483
D(H) 3.13E-06 3.63E-06 0.861662 0.3968
D(O) -0.006564 0.012315 -0.532989 0.5986
D(CG) 0.012411 0.054292 0.228593 0.8210
D(BG) 0.349634 0.351662 0.994232 0.3293
D(FN) 0.338243 0.565353 0.598286 0.5548
T -0.233178 0.576338 -0.404585 0.6891
RES-F 6.23E-03 0.000104 2.597376 0.0254
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
45
R-squared 0.111739 Mean dependent var -0.241955
Adjusted R-squared -0.229899 S.D. dependent var 7.431376
S.E. of regression 8.241457 Akaike info criterion 7.298005
Sum squared resid 1765.962 Schwarz criterion 7.776926
Log likelihood -124.0131 F-statistic 0.327069
Durbin-Watson stat 2.897256 Prob(F-statistic) 0.965926
Author’s computations
To recap, the models (both single and simultaneous equation models) provide evidence
that suggest that there is a bi-directional relationship between economic growth and FDI
inflows to Nigeria. Thus, as FDI encourages growth, more growth also encourages more
FDI, hence there is a kind of positive-feed-back relationship between FDI and economic
growth in Nigeria. Our results are different from that of Ayanwale (2007) in that we
identify four core determinants of FDI in Nigeria (openness, global rate of return on
investment, government consumption and the exchange rate) while Ayanwale identified
market size, infrastructure development and stable macroeconomic policies as core
determinants.
In summary, the empirical analyses conducted in this study reveal that there exists a bi-
directional relationship between FDI and economic growth in Nigeria. In other words,
as the economy improves, more FDI is attracted, and as more FDI flows into Nigeria,
the economy continues to grow. Thus, there is positive feedback relationship between
FDI and economic growth. The implication of these findings for future empirical
research is that empirical studies on FDI and growth in Nigeria should necessarily
account for the simultaneity bias, or endogeneity between FDI and Economic growth in
Nigeria.
POLICY DISCUSSION AND CONCLUSION
In summary, the models (both single and simultaneous equation models) provide
evidence that suggest that there is a bi-directional relationship between economic growth
and FDI inflows to Nigeria. Thus, as FDI encourages growth, more growth also
encourages more FDI, hence there is a kind of positive feedback effect in the
relationship between FDI and economic growth in Nigeria. These results have far
reaching implications for policy making in Nigeria. Specifically, the policy implications
are as follows.
I. Because of the crowding-out nature of government size in the economy, the
results suggest the reduction in government size in the economy. This is better achieved
through privatization and down-sizing of most government owned enterprises in the
Vol. 11, No.2 Journal of Monetary and Economic Integration
46
country. This will engender competition and greater efficiency. However, caution should
be exercised to ensure that the necessary conditions for privatization and down-sizing are
in place so as to avoid industrial actions, and the failure experienced during the first
privatization exercise in 1988. Government needs to provide the legal and administrative
framework for effective privatization. More importantly, there is the need to ensure
transparency in the exercise.
II. The results equally suggest the need to increase the degree of openness for
greater growth performance. Undoubtedly, development policies that are aimed at
ensuring greater private (domestic and foreign) participation in the economy will lead to
increase in the level of openness. This tends to buttress the argument that the economy
needs to be opened up through increased private participation. For example, foreign
investors participating in the debt conversion programme could be encouraged to direct
their investments to projects that significantly increase production capacity, incorporate
new technologies in the tradable sectors, and improve the country‟s infrastructure base.
III. Further, the negative sign of financial development possibly suggests the need
to stem the problem of capital flight and deepen the process of financial intermediation
in the country. Steps to level the legal and administrative playing field for domestic
investors and to promote a stable macroeconomic environment could contribute to
stemming capital flight. Policy makers therefore have the task to keep legitimate private
capital at home by encouraging domestic investment.
IV. Policies to encourage private holders of external assets to repatriate their capital
should be implemented. These possibly might include tax amnesties and raising the
domestic interest rate. It needs to be pointed out, however, that these policies could have
adverse effects on already weak private sector in the economy, but then, it will intensify
the flows of FDI into the domestic economy.
V. The findings on human capital points to the need for Nigeria to follow an
educational policy that would further raise the stock of human capital, especially at the
tertiary levels. This will aid faster technology diffusion and reduce the extent of capital
flight since intermediate and senior staff will no longer be foreigners, but nationals who
will retain their profits and incomes in the domestic economy.
On a general note, policies that require reducing political risk, ensuring property
rights, and policies that bolster growth in market size, as well as wage moderation (ideally
lowering real wages of political office holders), lowering corporate tax rates, and ensuring
full integration of the Nigerian economy into the world economy will go a long way in
reinforcing the positive feedback relationship between FDI and economic growth in
Nigeria.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
47
REFERENCES
Adelegan, J. O. (2000). Foreign direct investment and economic growth in Nigeria: A
seemingly unrelated model. African Review of Money, Finance and Banking,
Supplementary Issue of Savings and Development, pp. 5-25, Milan , Italy.
Ajayi, S. I. (2006). The determinants of foreign direct investment in Africa: A survey of
the evidence. In: Ajayi S.I. (Ed.) Foreign Direct Investment in Sub-Saharan Africa:
Origins, Targets, and Potential, AERC, Nairobi.
Akinlo, A. E. (2004). Foreign direct investment and growth in Nigeria: an empirical
investigation. Journal of Policy Modelling. 26, 627-639.
Alfaro, L., Chanda, A, Kalemli-Ozcan, S. and Sayek, S. (2004). FDI and Economic
Growth: The Role of Local Financial Markets. Journal of International Economics, 64
(1): 89-112.
Aluko, S. A. (1961). Financing economic development in Nigeria. The Nigerian Journal of
Economic and Social Studies. 3(1), 39-67.
Ariyo, A. (1998). Investment and Nigeria‟s economic growth. . In Investment in the Growth
Process: Proceedings of the Nigerian Economic Society Conference 1998, pp. 329–49.
Ibadan, Nigeria
Asiedu, E. (2002). On the determinants of foreign direct investment to developing
countries: Is Africa different. World Development, 30 (January): 107–19.
Asiedu, E. (2003). Foreign direct investment to Africa: the role of government policy,
governance and political instability. Working Paper, University of Kansas.
Asiedu, E. (2004). Policy reform and foreign direct investment to Africa: Absolute
progress but relative decline. Development Policy Review, 22(1): 41–8.
Ayanwale, A.B. (2007). FDI and economic growth: evidence from Nigeria. AERC
Research Paper 165, African Economic Research Consortium, Nairobi.
Ayanwale, A.B., and Bamire, A.S. (2001). The influence of FDI on firm level productivity
of Nigeria‟s Agro-Allied sector. Final Report Presented to the African Economic
Research Consortium, Nairobi.
Balasubramanyam, V. N., Salisu, M. A. and Sapsford, D. (1996). Foreign Direct
Investment and Growth in EP and IS Countries‟. Economic Journal, 106 (434): 92-
105.
Bator, F. M. (1958). The anatomy of market failure, Quarterly Journal of Economics, 12, 351-
379.
Baumann, J. G. (1975). Merger theory, property rights and the pattern of U.S Direct
Investment in Canada. Weltwirtschaftliches Archiv 7, 676-698.
Bende-Nadembe, A., Ford, J., Santoso, B. and Sen, S. (2003). The interaction between
FDI, output and the spillover variables: Cointegration and VAR analysisfor
APEC, 1996-1999. Applied Economics Letters, 10(3) 165-172.
Bennell, Paul. (1995). British manufacturing investment in Sub-Saharan Africa:
Corporate responses during structural adjustment. The Journal of Development
Studies, 32(2). 186-209.
Vol. 11, No.2 Journal of Monetary and Economic Integration
48
Bergsten, C. F., Horst, T., and Moran T. H. (1978). American Multinationals and American
Interests. The Brookings Institute, Washington D.C.
Blonigen, B., and Wang, M. (2005). Inappropriate pooling of wealthy and poor countries
in empirical FDI studies. In: Moran, T.H., E. Graham, and M. Blomstrom (Eds.).
Does Foreign Direct Investment Promote Development? Institute of International
Economics Press, Washington, D.C.
Borensztein, E., Gregorio, J. and Lee, J. (1998). How does foreign direct investment
affect economic growth? Journal of International Economics. 45 (1) 115-135.
Brown, C.V. (1962). External economies and economic development. The Nigerian Journal
of Economic and Social Studies, 4(1): 16–22.
Calvet, A. L. (1980). Market and Hierarchies: Towards a Theory of International Business.
Doctoral Dissertation, Sloan School of Management, M.I.T.
Carkovic, M., Levine, R., (2005). Does foreign direct investment accelerate economic
growth? In: Moran, T.H., Graham, E.M., Blomstr ¨om, M. (Eds.), Does Foreign
Direct Investment Promote Development? Institute of International Economics
Press, Washington, DC.
Caves, R. E. (1971). International cooperations: the industrial economics of foreign
investment. Economica, February, pp. 1-27.
CBN (2009). CBN Statistical Bulletin, Special Anniversary Edition, CBN, Abuja.
Chenery, H. B. and A. Stout (1966). Foreign Assistance and Economic Development.
American Economic Review Vol. 55 pp.679-733.
Chuku, C. A. (2009). Measuring the effects of monetary policy innovations in Nigeria: A
structural vector autoregressive (SVAR) approach. African Journal of Accounting,
Economics, Finance and Banking Research, 5 (5); 112-129
De Mello, L. R. (1997). Foreign direct investment in developing countries and growth: a
selective survey. Journal of Development Studies 34, 1-34.
Durham, B. (2004) Absorptive capacity and the effects of FDI and equity foreign
portfolio investment on economic growth. European Economic Review 48, 285- 306.
Edozien, E. G. (1968). Linkages, direct foreign investment and Nigeria‟s economic
development. The Nigerian Journal of Economic and Social Studies. 10(2); 119-203.
Ekpo, A. H. (1995) Foreign direct investment: evidence from time series data. CBN
Economic and Financial Review 35 (1) 59-78.
Elbadawi, I. and Francis, M. (1997). Regional integration and FDI in Sub-Saharan
Africa”. AERC Working Paper. African Economic Research Consortium, Nairobi.
Fedderke, J. W. (2002). The virtuous imperative: modelling capital flow in the presence
of non-linearity. Economic Modelling 19, 445-461.
Fedderke, J.W., and Romm, A. T. (2006). Growth impact and determinants of foreign
direct investments in South Africa, 1956-2003. Economic Modelling 23, 738-760.
Findlay, Ronald (1978). Relative Backwardness, direct foreign investment, and the
transfer of technology: A simple dynamic model. Quarterly Journal of Economics, 42,
371-393.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
49
Gastanaga, Victor, Jeffrey B. Nugent and Bistra Pashamova. (1998). Host country
reforms and FDI inflows: How much difference do they make? World Development,
26(7): 725–54.
Ghatak, A., and F. Halicioglu (2006). Foreign direct investment and economic growth:
Some evidence from across the World. MPRA Paper No. 3563.
Graham, E. M. (1974). Oligopolistic Imitation and European Direct Investment in the
United States. Doctoral Dissertation, Harvard Business School.
Greene, W. (2003). Econometric Analysis. Prentice Hall, New York.
Gujarati, D., and Porter, A. (2009) Basic Econometrics, McGraw-Hill, New York.
Hermes, N. and R. Lensink (2000). Foreign direct investment, financial development and
economic growth. SOM Research Report OOE27, University of Groningen, The
Netherlands.
Hsiao, C. and Y. Shen (2003). Foreign direct investment and economic growth: the
importance of institutions and urbanization. Economic Development and Cultural
Change. 51, 883-896.
Hymer, S. H. (1970). The International Operations of National Firms: A Study of Foreign Direct
Investment. Cambridge, MA MIT.
Kindleberger, C. P. (1969) American Business Abroad: Six Lectures on Direct Investment. Yale
University Press, New Haven.
Knickerbocker, F. T. (1974). Oligopolistic Reaction and Multinational Enterprise. Cambridge,
MA: Harvard Business School Division of Research.
Lehmann, A. (1999). Country risks and the investment activity of U.S. multinationals in
developing countries”. IMF Working Paper No. 133. International Monetary Fund,
Washington, D.C
Li, X and X. Liu (2005). Foreign direct investment and economic growth: An
increasingly endogenous relationship. World Development 33(3) 393-407.
Lim, Ewe-Ghee (2001). Determinants of and the relation between, foreign direct
investment and growth: a summary of recent literature. IMF Working Paper 175.
International Monetary Fund, Washington D.C.
Morisset, J. (2000). Foreign direct investment in Africa: Policies matter. Transnational
Corporations, 9(2): 107–25.
Obinna, O.E. 1983. “Diversification of Nigeria‟s external finances through strategic
foreign direct investment”. Nigerian Economic Society Annual Conference Proceedings,
Jos, 13-16th May.
Obwona, Marios B. (2004). Foreign direct investment in Africa. In Financing Pro-Poor
Growth:AERC Senior Policy Seminar VI, Kampala, Uganda, 2–4 March 2004 – Seminar
Papers, pp. 60–95. Nairobi: African Economic Research Consortium.
Odozi, V.A (1995). An overview of foreign investment in Nigeria1960-1995. Occasional
Paper No. 11 Research Department, Central Bank of Nigeria.
Oseghale, B. D., and Amonkhienan, E. E. (1987). Foreign debt, oil export, direct foreign
investment (1960-1984). The Nigerian Journal of Economic and Social Studies. 29(3)
359-380.
Vol. 11, No.2 Journal of Monetary and Economic Integration
50
Oyinlola, O. 1995. “External capital and economic development in Nigeria (1970–
1991)”. The Nigerian Journal of Economic and Social Studies, 37(2&3): 205–22.
Pigato, Miria. 2001. “The foreign direct investment environment in Africa”. Africa Region
Working Paper Series No. 15. The World Bank, Washington, D.C.
Ragazzi, G. (1973). Theories of the determinants of foreign direct investments, IMF Staff
Papers, July, pp. 471-498.
Ramasamy, Bala. (1999). FDI under uncertainty: Lessons from Malaysia”. Conference
Proceedings, International Conference on the Challenges of Globalization,
Thammassat University, Bangkok, Thailand, 21–22 October.
Ramirez, M. D. (2000) Foreign direct investment in Mexico: A cointegration analysis.
Journal of Development Studies 37, 138-162.
Read, R. (2007). Foreign direct investment in small island developing states, UNU-
WIDER, Research Paper No. 2007/28.
Reis, A. (2001). On the welfare effects of foreign investment. Journal of International
Economics 54, 411-427.
Romer, D. (2001). Advanced Macroeconomics. McGraw-Hill Irwin. New York.
Romer, P.M. (1993). Idea gaps and object gaps in economic development. Journal
ofMonetary Economics. 32, 215-234.
Ruxanda, G., and Muraru, A. (2010). FDI and economic growth: evidence from
simultaneous equation models. Romanian Journal of Economic Forecasting (1) 45-57.
Schneider, F. and B.S. Frey. (1985). Economic and political determinants of foreign
direct investment. World Development, 13(2): 171–75.
Schoeman, N.J., Clausen, R. Z. and de-Wet, T.J.. (2000). Foreign direct investment flows
and fiscal discipline in South Africa. South African Journal of Economics and
Management Science, 3(2): 235–44.
Vernon, R. (1966). International investment and international trade in the product cycle.
Quarterly Journal of Economics, May, pp. 223-247.
Vu, Tam-Bang, and . Noy, I (2009). Sectoral analysis of foreign direct investment and
growth in the developed countries. Journal of International Financial Markets,
Institutions and Money. 19, 402-413.
Zhang, K.H. (2001). “Does foreign direct investment promote economic growth?
Evidence from East Asia and Latin America. Contemporary Economics, 5, 217-312
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
51
APPENDIX
DATA SOURCES, CONSTRUCTION AND DESCRIPTION
I. Real Output (denoted by Y): this variable is measured by the real gross
domestic product at 1990 constant basic prices. We obtain the series from
the 2009 Edition of the Central Bank of Nigeria‟s Statistical Bulletin. We
further transform the variable to their natural logarithm.
II. Private Capital Stock (denoted by Kp): this variable is measured by the level
of private investments in Nigeria. We obtain the series from the Statistical
Bulletin of the Central Bank of Nigeria (2009). We also take the natural
logarithm of the series in other to express it as an index.
III. Stock of Foreign Investment (denoted by F): this variable is measured by
the inflows of foreign direct investments in Nigeria. It is obtained from the
current accounts section in the Balance of Payments accounts of Nigeria.
We also convert this variable to their natural logarithm in order to have
them indexed.
IV. Human Capital (denoted by H): this variable measures the importance of
human capital development on economic growth. We use education as our
proxy for human development, and we measure the variable by finding the
ratio of secondary and tertiary (universities, polytechnics and colleges of
education) enrolment in the total population. Adawo (2011), Anyanwale
(2007) and Akinlo (2006) have also measured human capital in similar ways.
We collect the required data series from publications of the Federal Ministry
of Education, National Bureau of Statistics and UNESCO.
V. Openness (denoted by O): we measure the openness of Nigeria as an FDI
host economy by taking the ratio of trade to GDP. In other words,
openness is measured as the ratio of (Import + export)/ GDP. The data sets
are obtained from the CBN Statistical Bulletin, 2009 Issue.
VI. Government Size (denoted by Cg): this variable is measured as the ratio of
government expenditure to GDP. It is expected that the higher the
government expenditure in the economy ceteris paribus, the higher will be
the level of growth. This is because government consumption will increase
aggregate demand and therefore, stimulate supply. This variable is obtained
from the Statistical Bulletin of the Central bank of Nigeria.
Vol. 11, No.2 Journal of Monetary and Economic Integration
52
VII. Savings ratio (denoted by Sn): this variable is introduced in the simultaneous
equation models, following the works of Ruxanda and Muraru (2010), who
use it to measure the efficiency of absorption of FDI into the domestic
economy.
VIII. Gross Capital Formation (denoted by M): this variable is used to capture the
rate of capital accumulation in the country. It is measured as the ration of
gross capital formation to GDP. The series are obtained from the World
Bank‟s World Development Indicators.
IX. Budget Balance (denoted by Bg): this variable measures the fiscal climate in
Nigeria. It is measured as the ratio of budget balance (deficit and surpluses)
to GDP. Akinlo (2006) also measured it in a similar way. We obtain data on
fiscal balances from the Federal Ministry of Finance and augment it with
that from the Statistical bulletin of the Central bank of Nigeria.
X. Financial Development (denoted by Fn): this variable measures the effect of
financial development on economic growth in Nigeria. It is measured by
taking the ratio of money supply, broadly defined as M2 to GDP. As usual,
the values of these series are obtained from the Statistical Bulletin of the
Central bank of Nigeria.
XI. Level of Infrastructure (denoted by I): in the literature, most studies in
advanced economies have used the number of telephones per 1000
population as a proxy for infrastructure. This measure of infrastructure is
however constricted by many other factors, thereby rendering it
inappropriate for a developing economy like Nigeria. Hence, we follow
Anyanwale (2007) by measuring the level of infrastructure development by
using the electricity consumption per capita. This series are obtained from
the World Development Indicators (WDI) of the world Bank. Freely
available at www.worldbank.org/wdi
XII. Return on Investment (denoted by R): we measure this variable by
following Ekpo (1995) who measured the return on investment in the rest
of the world, using long-term interest rate of US bonds. We obtain this
series from the Statistical Bulletin of the National Bureau of Economic
Research (NBER).
XIII. Per Capita Income (denoted by YP): we measure per capita income by
using the GDP per capita as measured by the purchasing power parity
approach. The series are obtained from the World development Indicators
of the World Bank.
Vol. 9, No.2 Okon J. Umoh, Augustine O. Jacob and Chuku A. Chuku
53
XIV. Inflation (denoted by N): we follow the norm in econometric research by
measuring the rate of inflation using the log first difference of the consumer
price index (CPI). The data sets are obtained from the Statistical Bulletin of
the Central Bank of Nigeria.
XV. Labour Force (denoted by L): this variable is measured by the series of
economically active labour force. That is, the official statistics of labour
employed in the formal sector in Nigeria4. The data is obtained from the
Federal Ministry of Labour and Productivity and supplemented with
information from the National Bureau of Statistics, i.e. the Annual Abstract
of Statistics (various issues).
XVI. Real Effective Exchange Rate (denoted by E): The real effective exchange
rate is obtained by deflating the nominal exchange rate by an index of the
exchange rate between Nigeria and three of its major trading partners.
(America, China and Europe).
XVII. Period Adjustment Dummy (denoted by D): we differentiate between the
periods of the structural adjustment programme and the periods of no
adjustment by including a dummy variable. The dummy is such that during
the periods of adjustment, a value of 1 is assigned to the dummy, and during
periods of no adjustment, 0 is assigned.
XVIII. Trend Adjustment Dummy (denoted by T): since our analysis spans
between several years, we introduce a time adjustment dummy to capture
secular trends in output during the period under review.
4 We acknowledge that this may be an underestimated value of the economically active labour force since those in the
large informal sector are not capture. However, since there are no official records about the activities of labour outside
the formal sector. We improvise with the information from the formal sector.
