Foreign direct investment and domestic investment: Do oil sectors matter? Evidence from oil-exporting Gulf Cooperation Council economies

Accepted Manuscript Title: Foreign direct investment and domestic investment: Do oil sectors matter? Evidence from oil-exporting Gulf Cooperation Council economies Author: Mohamed Elheddad PII: DOI: Reference:

S0148-6195(18)30036-5 https://doi.org/10.1016/j.jeconbus.2018.11.001 JEB 5828

To appear in:

Journal of Economics and Business

Received date: Revised date: Accepted date:

8 January 2018 12 November 2018 13 November 2018

Please cite this article as: Elheddad M, Foreign direct investment and domestic investment: Do oil sectors matter? Evidence from oil-exporting Gulf Cooperation Council economies, Journal of Economics and Business (2018), https://doi.org/10.1016/j.jeconbus.2018.11.001 This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.

Foreign direct investment and domestic investment: Do oil sectors matter? Evidence from oil-exporting Gulf Cooperation Council economies

IP T

Mohamed Elheddad [email protected]

SC R

Corresponding author: Mohamed Elheddad, University of Hull, 34 Sidmouth Street, HU5 2LB, Hull

Highlights:

U

 Effects of aggregate and disaggregates FDI on domestic investments tested in GCC

N

 FDI inflows contribute to public investments but discourage private

A

investments

M

 FDI outflows promote private investments but hinder public investments

Non-Oil FDI has an ambiguous effect on domestic investments

CC E

Abstract

PT



ED

 Oil FDI inflows yield a positive effect on public investment

This article investigates the impact of sectoral foreign direct investment (FDI) on public and private domestic investment (DI) using a unique dataset of greenfield FDI inflows to six oil-

A

exporting economies of the Gulf Cooperation Council (GCC) during 2003-2013. After controlling for some econometric issues and macroeconomic determinants of DI, there are two main findings. First, aggregate estimations show that FDI inflows contribute significantly to public DI but discourage private DI. Conversely, FDI outflows promote private domestic economic activities but negatively affect public DI. Second, disaggregated data illustrates that 1

greenfield FDI inflows to the oil sector yield a significant and positive effect on public DI; a 1% increase in oil FDI leads to about a 0.6% improvement in public DI. Non-oil FDI has an ambiguous effect on DI. All the results are robust.

Keywords: FDI, Public domestic investment, Private domestic investment, Panel data,

IP T

Endogeneity, Oil, GCC

A

CC E

PT

ED

M

A

N

U

SC R

JEL codes: F21; F23; F62; E22; C23; C26

2

1. Introduction Foreign direct investment (FDI), through providing new technology and investment accrual, is seen as one of the significant benefits of globalisation (Morrissey & Udomkerdmongkol, 2012; Farla et al., 2016). FDI has also become a significant source of finance, larger than official

IP T

development aid (Ayyagari & Kosová, 2010). Most developing countries have therefore introduced new policies and regulations to attract foreign investors, including preferential tax

SC R

systems and preferential loans (Farla et al., 2016).

The main concern is whether FDI crowds-in or crowds-out domestic investment (DI), whether public or private. This is a very important issue, but there are few empirical studies on the

U

effects of sectoral FDI on DI. Previous literature on the relationship between FDI and DI has

N

arrived at contradictory findings. While some studies have found that FDI stimulates DI

M

(Morrissey & Udomkerdmongkol, 2012)

A

(Ayyagari & Kosová, 2010; Farla et al., 2016), others have concluded that FDI replaces DI

ED

Based on the empirical work, there are two central hypotheses – the influence of FDI could be either positive or negative. FDI inflows could adversely affect local firms through a

PT

competition effect (Haddad & Harrison, 1993; Aitken & Harrison, 1999) or, alternatively, foreign firms could generate positive spillovers by providing new products or introducing new

CC E

technologies (Teece, 1977; Smarzynska Javorcik, 2004). Furthermore, FDI outflows may enhance DI by combining home production and foreign production to reduce costs and increase

A

the returns to domestic production (Desai et al., 2005). Alternatively, FDI outflows may substitute foreign activities for DI, with firms shifting their production factors abroad (Stevens & Lipsey, 1992). The purpose of this paper is, first, to re-examine Feldstein’s (1995) hypothesis that outbound FDI reduces DI, utilising data for six oil-abundant countries in the Gulf Cooperation Council 3

(GCC). This paper then investigates whether the concentration of FDI inflows in the oil sector promotes or deters DI by examining the impacts of sectoral FDI inflows on both public and private DI. The GCC, with its significant geo-economic position, attracts a vast amount of FDI inflows.

IP T

During a booming oil price period (2000-2010), FDI inflows jumped by 2,533%, outpacing the world average by 108% during the same period (Toone, 2012). These foreign investments were concentrated in the oil sector, with more than 60% of total FDI inflows in the oil industry

SC R

between 2003 and 2013, while non-oil sectors (including manufacturing and services) received about 39% of total FDI inflows (Financial Times, 2016).

U

Further, the GCC economies are dominated by a public sector in which wages are high, the

N

work environment is easy, and jobs are held by local employees (Hertog, 2012). In contrast,

A

the private sector offers low quality of employment and the jobs are held by foreign workers.

M

Private investments are also managed and owned by political elites, such as princesses, and

ED

financed mainly by the hydrocarbon industry (Hertog, 2013). Improving the quality of FDI in oil-producing economies may generate positive spillovers into DI and help policymakers to

PT

adopt an effective diversification policy.

Although the flood of FDI inflows to and outflows from developing countries may raise a

CC E

significant question regarding their efficiency and impacts on DI, there is lack of literature analysing FDI inflows to different sectors in relation to private and public DI in the host country.

A

Government and policymakers need to evaluate how DI is affected by FDI; understanding these effects may help such countries, and in particular those in the GCC, to introduce proper policies that promote diversification. This study contributes to the literature in three ways. First, it investigates the sectoral effects of FDI on local private and public investments, which will help policymakers to increase the 4

quality of FDI in the host country. Previous papers have focused on the aggregate level of FDI (Feldstein, 1995; Borensztein et al., 1998; Agosin & Machado, 2005; Desai et al., 2005; Herzer, 2008; Ghassan & Alhajhoj, 2012). This study, in contrast, distinguishes itself by using a unique, detailed dataset on greenfield FDI inflows to several sectors. This kind of data allows examination of which kinds of FDI complements or replaces DI. Second, this study considers

IP T

the issue of expected reverse causality between FDI and DI, and the duality between saving

and DI (Feldstein & Horioka, 1980a; Feldstein, 1995). Although there are some studies that

SC R

control for endogeneity by applying generalised method of moments ( GMM) estimator (AlSadiq, 2013), this study is the first empirical paper to apply the instrumental variable approach

U

to estimating the FDI-DI relationship. Third, this article, to the best of my knowledge, is the

N

first study to focus on the oil-exporting economies of the GCC countries using disaggregated

A

data.

M

This paper is structured as follows: after this introduction, section 2 reviews the previous literature on the relationship between FDI and DI. Section 3 provides a theoretical framework

ED

for the FDI-DI relationship, and section 4 discusses data and methodology. Section 5 presents

A

CC E

implications.

PT

the empirical results and, finally, section 6 provides the main conclusions and policy

5

2. FDI and domestic investment: Empirical literature The existing literature on the association between FDI and DI is far from conclusive. Some papers have found that FDI enhances DI, but others have reached the opposite conclusion, that FDI reduces DI. Feldstein (1995) is among the most significant and cited works in favour of the impact of FDI

IP T

on DI. Utilising a macro-level dataset for OECD countries during the 1970s and 1980s,

Feldstein (1995) found that higher outward FDI reduced DI, while FDI inflows had a positive

SC R

impact on local investments after controlling for several macroeconomic determinants of DI.

Andersen and Hainaut (1998), using a dataset for the 1960s and 1990s, reported that FDI

U

outflows tended to reduce DI in the United States, Japan, Germany, and the United Kingdom.

N

There are several mechanisms that explain the negative relationship between outward FDI and

A

local investments. Among the earliest ideas is the substitution of foreign activities for DI –

M

firms shift their production factors abroad (Stevens & Lipsey, 1992).

ED

Taking the opposite view, firm-level studies have argued that outward FDI could promote DI. Outward FDI enables corporations to enter new markets, have access to intermediate goods at

PT

lower prices, and gain access to foreign technologies. Based on this process, outboundinvesting firms increase their competitiveness by combining home production with foreign

CC E

production. Stevens and Lipsey (1992), using firm-level data for seven American firms over periods of 16 to 20 years, found that there is a significant and positive association between

A

outward FDI and DI.

Similar results have been found by other authors; for instance, Desai et al. (2005) compared macro models to micro models, first replicating Feldstein’s estimates using a broader sample of countries during the 1980s and 1990s, using OECD-country aggregated data, then using firm-level data for US multinational corporations (MNCs). Their empirical study found two

6

different results. Using macro-level estimations, FDI had a displacement effect on DI, confirming the results of Feldstein (1995); however, higher broad investment by American firms was positively associated with DI through a combination of home production and foreign production that produced final products at lower cost. Considering the short- and long-term impacts, Herzer and Schrooten (2008) found that outward FDI had a positive effect, in the short

Broad investment complemented DI only in the short term in Germany.

IP T

term, on German DI, while DI was negatively correlated with FDI outflows in the long term.

SC R

Along the same lines, Hejazi and Pauly (2003) found that the effect of outward FDI by

Canadian MNCs varied depending on the investment partner. Using a sectoral dataset over the

U

period from 1984 to 1995, they found that Canadian outward investment in the US

N

complemented DI in Canada, while investment in the rest of the world had a negative effect.

A

Considering the origins of investors, Ni et al. (2017), using a Vietnamese firm-level dataset for

M

2001-2011, investigated whether the origins of foreign investors generated different effects on

ED

DI in Vietnam. The main findings of the study were that Asian foreign investors contributed positively to local firms, but the effects of foreign investments from North America were not

PT

significant. With regard to Asian FDI, the study concluded that investments from China and Taiwan had the most significant impacts on local firms in Vietnam.

CC E

Focusing on inter-industry interactions between MNCs and domestic markets in eastern European countries during 2001-2007, Hanousek et al. (2017) found that FDI raised demand

A

for intermediate goods, but domestic suppliers benefitted more from FDI inflows only through positive shocks. Thereafter, they faced a crowding-out effect by MNCs when the larger companies entered upstream sectors. Surprisingly, very few studies have tried to investigate the impacts, on a sectoral level, of FDI on domestic economic activities. Only one empirical work, to the best of the author’s 7

knowledge, has examined the impacts of sectoral FDI inflow on domestic entrepreneurship. Using disaggregated data on FDI inflows into 96 countries during 2004-2012, Doytch (2016) found that service and mining FDI improved DI, and particularly financial FDI crowding in local activities. However, FDI in manufacturing yielded a negative impact.

IP T

Although previous studies have used several approaches and different datasets, they did not control for one serious methodological issue: the endogeneity of outbound FDI. The

endogeneity issue occurs when there is reverse causation between DI and FDI, and between

SC R

saving ratio and DI. Feldstein (1995) stated, ‘[a] country that offers a good environment for DI is also likely to attract more inbound FDI and may also experience less outbound FDI’

U

(Feldstein, 1995:55). Feldstein failed to find a proper instrumental variable and proposed some

N

likely instruments, but could not find data on those variables. Instead, he introduced more

A

explanatory variables, which may correlate with both FDI and DI. Al-Sadiq (2013) and Doytch

M

(2016) are exceptional empirical works that considered the issue of endogeneity between FDI and DI using system GMM. Al-Sadiq (2013) applied panel data analysis for 121 developing

ED

countries during 1990-2010 and found that FDI outflows had negative impacts on DI.

PT

The current study differs from the previous literature by focusing on a specific region – GCC countries – during 2003-2013 using sectoral level data on FDI inflows, and it differs from Al-

CC E

Sadiq (2013) and Doytch (2016) by using an instrumental variables estimator to control for

A

endogeneity.

8

3. Theoretical framework Agosin and Machado (2005) proposed a theoretical framework for investigating the crowdingout (CO) or crowding-in (CI) effects of FDI on DI. This study modifies Agosin and Machado’s (2005) model to fit the sample in this study’s dataset. It is assumed that FDI is an exogenous variable from the viewpoint of the host economy. In

IP T

this case, FDI depends mainly on global macroeconomic factors and the strategies of MNCs

(Agosin & Machado, 2005). The theoretical analysis starts from the idea that total investment

Therefore, the investment equation can be written as follows:

U

𝐼𝑡 = 𝐼𝑑,𝑡 + 𝐼𝑓,𝑡

SC R

in one economy is a summation of DI and foreign firms’ investment.

(1)

A

N

where 𝐼𝑡 is total investment at time t, 𝐼𝑑,𝑡 is DI at time t and 𝐼𝑓,𝑡 is FDI in this economy at time

M

t.

DI can be defined as a stock adjustment variable reacting to variances between desired and

(2)

PT

∗ 𝐼𝑑,𝑡 = 𝜏(𝐾𝑑,𝑡 − 𝐾𝑑,𝑡 )

ED

actual capital stock. Therefore, the DI equation can be written as:

CC E

∗ where 𝐾𝑑,𝑡 refers to the capital stock desired by domestic firms and 𝜏 > 1.

Following the neo-classical model of investment, desired capital relies on the expected growth rate 𝐺 𝑒 and the variation between expected output 𝑦 and actual output 𝑌. The desired capital

A

equation can then be written as: ∗ 𝐾𝑑,𝑡 = 𝛾0 + 𝛾1 𝐺𝑡+𝑡 + 𝛾2 𝑦𝑡

(3)

Substituting equation (3) into equation (2) gives 𝐼𝑑,𝑡 = 𝜏{(𝛾0 + 𝛾1 𝐺𝑡+𝑡 + 𝛾2 𝑦𝑡 ) − 𝐾𝑑,𝑡 }

(4) 9

= 𝜏𝛾0 + 𝜏𝛾1 𝐺𝑡+𝑡 + 𝜏𝛾2 𝑦𝑡 − 𝜏𝐾𝑑,𝑡

(5)

= 𝜃0 + 𝜃1 𝐺𝑡+𝑡 + 𝜃2 𝑦𝑡 + 𝜃3 𝐾𝑑,𝑡

(6)

where 𝜃0 = 𝜏𝛾0 , 𝜃1 = 𝜏𝛾1 , 𝜃2 = 𝜏𝛾2 and 𝜃3 = − 𝜏; the FDI equation is then

IP T

𝐼𝑓,𝑡 = ∅0 𝐹𝐷𝐼𝑖𝑛𝑓𝑙𝑜𝑤𝑡 + ∅1 𝐹𝐷𝐼𝑜𝑢𝑡𝑓𝑙𝑜𝑤𝑡

(7)

SC R

Putting them all together in equation (1) gives 𝜋0 + 𝜋1 𝐺𝑡+𝑡 + 𝜋2 𝑦𝑡 + 𝜋3 𝐾𝑑,𝑡 + 𝜋4 𝐹𝐷𝐼𝑖𝑛𝑓𝑙𝑜𝑤𝑡 + 𝜋5 𝐹𝐷𝐼𝑜𝑢𝑡𝑓𝑙𝑜𝑤

(8)

U

CO or CI effects depend mainly on 𝜋4 and 𝜋5 . If 𝜋4 > 0 and 𝜋5 > 0, this indicates a CI effect,

A

CC E

PT

ED

M

A

N

whereas if 𝜋4 < 0 and 𝜋5 < 0, it is a CO effect.

10

4. Data and methodology 4.1. Data The data for this study covers a sample of six oil-exporting and -producing countries – Bahrain, Kuwait, Qatar, Oman, Saudi Arabia, and the United Arab Emirates – in the period from 2003 to 2013. The reason for choosing this period is the availability of data on FDI by sector for the

IP T

variables of interest. To examine the effect of FDI on specific domestic sectors, DI is divided into public and private investment. As a proxy for public DI, the approach of Feldstein (1995),

SC R

Borensztein et al. (1998), Hejazi and Pauly (2003), and Herzer and Schrooten (2008) is followed, with public investment measured by gross fixed capital formation (GFCF) as a share of GDP. According to the World Bank, GFCF includes land improvements (fences, ditches,

U

drains, and so on); plant, machinery, and equipment purchases; and the construction of roads,

N

railways, and the like, including schools, offices, hospitals, private residential dwellings, and

A

commercial and industrial buildings. According to the 1993 SNA, net acquisitions of valuables

M

are also considered capital formation. This variable is taken from the World Economic

ED

Indicators database.

There are many proxies for private DI, including gross capital formation in the private sector

PT

as a percentage of GDP and domestic credit issued to the private sector as a share of GDP. Due to data constraints, this study adopts the latter proxy. Domestic credit to the private sector refers

CC E

to financial resources provided to the private sector by financial corporations, such as through loans, purchases of nonequity securities, and trade credits and other accounts receivable that

A

establish a claim for repayment. For some countries, these claims include credit to public enterprises. The current study uses two different datasets as sources of data. First is aggregate FDI (inflow and outflow ‘stock’). The stock of FDI (inwards and outwards) is defined by United Nations Conference on Trade and Development (UNCTAD) as ‘[f]or associate and

11

subsidiary enterprises, it is the value of the share of their capital and reserves (including retained profits) attributable to the parent enterprise (this is equal to total assets minus total liabilities), plus the net indebtedness of the associate or subsidiary to the parent firm. For branches, it is the value of fixed assets and the value of current assets and investments, excluding amounts due from the parent, less liabilities to third parties’ (UNCTAD, 2018).

IP T

This variable is drawn from the UNCTAD database.

Second is disaggregated greenfield FDI inflows. The data on greenfield FDI is taken from

SC R

fDiMarkets, part of fDi Intelligence, itself part of the Financial Times Group. fDiMarkets has

been tracing and verifying individual cross-border greenfield investment projects since 2003

U

and is now a primary source of data for UNCTAD, the World Bank, and the Economist

N

Intelligence Unit (Toews & Vézina, 2017). The database offers information on the value of

A

investments and the estimated number of jobs created. The main advantage of this type of

M

investment data is that it is less affected by measurement problems (Canton & Solera, 2016).

ED

Data from fDiMarkets contains 35 sectors. The study data was constructed around two main sectors to achieve the goals of this paper. The resource sector of an economy is that which

PT

makes direct use of natural resources or exploits natural resources. This includes agriculture, forestry, fishing, mining, and oil. This study includes the FDI inflows to coal, oil, natural gas,

CC E

minerals, and metals as resource FDI. The non-resource sector refers to the secondary sector that produces manufactured goods and

A

the tertiary sector that provides services. This study includes warehousing, energy, building and construction, industrial semipro, automotive, ceramics, plastics, beverages, consumer discretionary, non-automotive, automotive components, engines, turbines, textiles, biotech, paper and printing, medical devices, business machines, consumer staples, industrial machinery, and electronic components. The service sector includes real estate, hotels and

12

tourism, financial services, communications, business services, transportation, software and IT, aerospace, leisure and entertainment, pharmacies, space and defence, and healthcare. This kind of detailed data allows checking of the mechanisms by which FDI inflows to specific sectors enhance or deter DI. This distinction is important to evaluate the quality of FDI and

IP T

helps policymakers to plan their future strategies. Following previous literature, this study controls for other macroeconomic variables to make

SC R

the estimations more robust. The inflation rate is measured by annual percentage change in

consumer price index (CPI) to capture the impact of macroeconomic stability. In addition, to

trade and growth rate, measured by the change in GDP.

U

capture the impact of economy size, we include trade openness as a measure of international

N

Furthermore, we adopt the saving ratio as a determinant of DI, following the Feldstein-Horioka

A

model (Feldstein & Horioka, 1980). All control variables are taken from the World Economic

M

Indicators (World Bank). Table 1-A shows the definitions and sources of data used in this paper,

CC E

PT

ED

and Table 1-B presents the descriptive statistics of all the selected variables.

Insert Table 1-A here Insert Table 1-B here

A

Before conducting any empirical investigation, it is important to check for time series properties, to avoid the issue of spurious regression and to make the results more robust. Therefore, a unit root test for panel data was applied first. The traditional Augmented DickyFuller (ADF) arguably suffers from issues related to its power in rejecting the null stationary series, particularly for short-spanned data. Recent studies have suggested that panel unit root

13

tests are more powerful than unit root tests based on individual time series (Al-Iriani, 2006). Hadri (2000) proposed a unit root test for panel data, and it has been argued that this test is suitable for a small sample of the panel (small N). Since the current sample is relatively small, for some countries, Hadri’s unit root test is applied to the panel. Table 2 shows the results of Hadri’s unit root test. It is clear that all variables have a unit root

IP T

at level because the null hypothesis – that all panels are stationary – is rejected at both the 1% and 5% levels of significance. However, these series become stationary after taking the first

SC R

difference. Thus, these variables are all integrated from the first order I(1).

The next step is to identify whether there is a long relationship among the selected variables.

U

This study uses Pedroni’s (2004) co-integration test. Perdoni (2004) developed a technique that

N

enables the use of a small sample of panels (N, number of cross-countries) and allows for

A

heterogeneity in the intercepts and slopes of the co-integrating equation. The results of the co-

M

integration test are presented in Table 3 and show that the null hypothesis of no co-integration

ED

is rejected at the 1% significance level. This confirms that all dependent variables and independent variables share a long-run relationship in the GCC countries. This relationship

Insert Table 2 here Insert Table 3 here

A

CC E

PT

must be investigated by testing the size and direction of the impact using panel data estimations.

14

4.2 Methodology and some econometric issues This study uses panel data analysis. Most of previous studies start with pooled Ordinary Least Squares (OLS) as a principal estimation. However, this method mixes time series and crosssection, an approach, which creates two major issues: failure to account for heterogeneity among panels (unobserved country-specific effects) and the expected endogeneity problem,

IP T

since FDI can affect DI through spillover effects but, at the same time, the level and quality of DI may attract MNCs – in other words, reverse causality. More incentives provided to domestic

SC R

firms could also motivate more FDI, creating a problem of omitted variables.

The fixed-effect model is therefore applied, which helps to mitigate the issue of heterogeneity but does not control the potential for endogeneity. Most empirical studies have used GMM

U

estimators to overcome this problem. There are many techniques for GMM, including Arellano

N

and Bond (1991), Blundell and Bond (1998), and recently Roodman (2009). It is argued that

A

GMM is appropriate for studies that have a longer cross-sectional (N) than time (T) series

M

(Roodman, 2009). GMM also suffers from the problem of instrument proliferation. This study

ED

therefore applies an instrumental variables estimator, which is suitable for a small sample (Cameron, 2010), specifically the instrumental variables proposed by Feldstein and Horioka

PT

(1980) and Feldstein (1995). We instrumented DI by population growth, retired people (over 65 years old) as a percentage of young people, and labour participation. These variables are

CC E

believed to respond simultaneously to changes in DI, consistent with Feldstein (1995). Furthermore, there is an expectation that domestic and foreign investments simultaneously

A

affect each other (endogeneity); for this purpose, the current study uses ease of doing business, freedom of investment, and trade freedom.

15

5. FDI and domestic investment: Empirical evidence and discussion Departing from an aggregate analysis of the impact of FDI (inflows and outflows) on DI (public and private), this study tries to re-produce Feldstein's (1995) model for six oil-dependent countries, using World Bank data. It then examines which sectoral FDI generates CI or CO effects on domestic economic activities using a two-sector model utilising data from

5.1. FDI and domestic investment: Aggregate-level analysis

IP T

fDiMarkets.

SC R

The main debate on the FDI-DI association is whether FDI crowds-in or crowds-out DI. This has been investigated in several studies (Feldstein & Horioka, 1980; Feldstein, 1995; Borensztein et al., 1998; Agosin & Machado, 2005; Desai et al., 2005; Al-Sadiq, 2013; Doytch,

N

This analysis is based on the following specifications:

U

2016). This section re-examines Feldstein’s hypothesis (1995) for oil-exporting economies.

A

𝐷𝐼𝑖,𝑡 = 𝛽0 + 𝛽1 𝐹𝐷𝐼𝑖𝑛𝑓𝑙𝑜𝑤𝑠𝑖,𝑡 + 𝛽2 𝐹𝐷𝐼𝑜𝑢𝑡𝑓𝑙𝑜𝑤𝑠𝑖,𝑡 + 𝛽3 𝑠𝑎𝑣𝑛𝑔 𝑟𝑎𝑡𝑖𝑜𝑖,𝑡 + 𝛽3 𝑋𝑖,𝑡 + 𝜀𝑖,𝑡 +

M

𝜗𝑖,𝑡

(9)

ED

where DI and FDI represent domestic investments and foreign direct investments, respectively, in country i at time t; saving ratio represents saving to GDP ratio; X is a vector of other

PT

determinants of domestic investments; 𝜀𝑖,𝑡 is a time-invariant unobserved heterogeneity term;

CC E

and 𝜗𝑖,𝑡 is the random error term. Now, the investigation turns to a specific area: GCC countries. In this area, the impacts of FDI

A

inflows into different sectors on private and public DI are investigated using the above specifications, but with DI separated into private and public. Tables 4, columns 1 and 2, report the results of the fixed effects model after controlling for heterogeneity. These results show that there is a non-significant positive effect of inbound FDI on public DI, but discouragement of private DI, in the GCC economies. A 1% increase in FDI

16

inflows to GCC countries leads to about a 0.00915% increase in public investments and a 0.092% decrease in private domestic activities. All these results are robust, even after controlling for other macroeconomic variables and significant at the 5% level. However, after performing the instrumental variables (IV) estimation, the negative effect of

IP T

FDI inflows on private investments becomes significant as the magnitude grows. A 1% rise in FDI inflows causes about a 0.358% reduction in the private DI. On the other hand, FDI inflows increase public DI by 0.32% for each 1% rise. These results offer evidence of a CO effect

SC R

between FDI inflows and private investments.

The negative association between FDI inflows and private DI can be explained by a

U

competition effect. MNCs are larger than domestic firms, and these foreign firms use very

N

advanced technology, which may not be available to domestic firms. (Haddad & Harrison,

A

1993; Aitken & Harrison, 1999). On the other hand, public investments in GCC countries

M

depend mainly on the oil sector to finance their activities; therefore, these investments benefit

ED

more from foreign firms’ activities through their profits and technologies. For that reason, the impacts of oil-oriented FDI on DI are later investigated. It can also be suggested that FDI is

environment.

PT

slow to engage with private investment because of the uncertain political and economic

CC E

Outflow FDI crowds-in private DI in GCC countries. Its positive effect ranges from 0.605 to 0.624 – private investment increases by 6.24% for each 10% increase in investment abroad.

A

These estimations are robust for several estimators. One possible explanation for this positive impact is that outward FDI permits corporations to import cheaper factors from foreign affiliates and to make exports of intermediate goods used by foreign affiliates. This means those firms make a combination of home production and foreign production to reduce costs and thereby raise the returns on domestic production. (Desai et al., 2005; Herzer, 2008)

17

However, outbound FDI yields an adverse effect on public DI of about 0.37, after controlling for unobserved variables and country-specific effects (fixed effects model). This indicates that, if there is increased outward investment of 10%, public DI drops by 3.7%. Moreover, the Instrumental Variables (IV) estimation illustrates larger negative impacts of

IP T

outward FDI on public investments compared with the Fixed Effects (FE) estimations. IV estimation (Table 4, column 4) shows that this negative effect becomes 0.62, which means outbound FDI leads to greater reduction in public investments. These results are expected

SC R

because, when firms intend to invest abroad, they move parts of production to other countries,

A

CC E

PT

ED

M

A

N

Insert Table 4 here

U

which causes a shortage of production factors in these countries (the origin of firms).

18

5.2. FDI and domestic investment: Sector-level analysis To examine what kind of FDI crowds-in or crowds-out public and private DI, in this section, FDI inflows are split into two main sectors: oil and non-oil. Given the importance of the oil boom as a key determinant of foreign investors’ decisions in GCC countries, it is worth investigating how the effects of FDI inflows differ between resource

IP T

and non-resource sectors. For this purpose, this study splits the FDI into two types: FDI in the

oil and non-oil industries. This kind of analysis shows whether FDI is leading or following the

SC R

distribution of host country production. If the sectoral distribution of FDI is noticeably different from the distribution of existing capital stock or of production, the contribution of FDI to capital formation is likely to be more positive than when the distribution of FDI follows the

N

U

existing sectoral distribution of capital stock (Agosin & Machado, 2005).

A

Table 5 shows the impact of oil and non-oil FDI on DI. All the results confirm that FDI into

insignificant effect on private DI.

M

extractive industries has a significant CI effect on public DI, but these investments have an

ED

Public DI gains about 0.6% when FDI inflows to the oil sector increase by 1%. This effect is

PT

stable and robust for the IV estimator, even after controlling for other macroeconomic variables. The reason behind this behaviour of FDI in the extractive industry is that the GCC is petroleum-

CC E

dependent. These countries invest massively in infrastructure and depend predominantly on only one source of revenue: oil. With that in mind, GCC economies attracted almost 50% of FDI inflow to the Middle East and North Africa (MENA) region during the 2000s, and these

A

investments were concentrated mainly in extractive industries. Therefore, it is expected that oil FDI would have a positive impact on the public sector. These results show that FDI in these economies is following the structure of production. Private DI is a relatively weak sector in the GCC because of high political and economic uncertainty, and less conducive regularity and institutional environments (World Bank, 2011; 19

IMF, 2016). Because of this, foreign firms do not participate much in private economic activities.

IP T

Insert Table 5 here

5.3. The impact of oil price shocks on domestic and foreign investments

SC R

To test the impact of positive oil prices on the variables used, this study applies the vector autoregression (VAR) model, which is valid as all the variables are stationary at first difference I(1)

and are co-integrated (Sahoo et al., 2014). In a VAR model, the coefficient cannot be explained

U

directly, and the innovation-accounting techniques have therefore been adopted, which consist

N

of impulse response functions (IRF). The IRFs inspect the relative effects of each variable on

A

other variables and display the response of each concerned variable in the linear system to a

M

shock from system variables. The main part of the VAR model is the optimal lag selection; the

ED

lag-length selection in the current VAR estimation is based on the Akaike information criterion (AIC), Schwarz information criterion (SIC), and Hannan-Quinn information criterion (HQC).

five years.

PT

The details of lag selection are presented in Table 6; the criteria agree on an appropriate lag of

CC E

Figure (1) presents the results of the IRFs of oil price on DI (public-private). It is clear that oil price has initial positive effects on public DI, but these effects become negative after the second

A

year. An increase in oil prices by one standard deviation leads to an increase in public investments for the next two years. Private DI is found to respond negatively to a shock in oil prices for the first two years. These responses then become positive after the third year. The latter result is in line with Hanousek et al. (2017); private investment for the GCC economies may benefit from future positive shocks in oil price.

20

Another interesting result is related to the responses of oil-based and non–oil related FDI inflows to oil prices. Figure (2 shows that FDI inflows to the oil sector respond positively to a one standard deviation increase in oil prices. However, this effect becomes negative after two

IP T

years, and vanishes after six years.

Insert Table 6 here

A

CC E

PT

ED

M

A

N

U

Insert Figure 2 here

SC R

Insert Figure 1 here

21

6. Conclusions and policy implications This paper provides novel empirical evidence of the effects of different types of FDI on the DI ratio in the oil-rich economies of the GCC countries. While there are a substantial number of empirical studies on the impacts of inflows and outflows of total FDI on total DI, to the best of the author’s knowledge, this is the first such study to investigate disaggregated FDI impacts on

IP T

private and public investments. Theoretically, the consequences of inward FDI on the host country’s DI ratio depend on the

SC R

motives for investing abroad. Because different types of FDI cannot be distinguished by sector, the effect of FDI on the home country’s DI becomes a debatable question. Therefore, the main

U

objective of this empirical paper is to test the association between inwards FDI in different

N

sectors and the country's rate of DI, using a unique dataset from six oil-producing countries,

A

over the period from 2003 to 2013, and a variety of different estimators to provide robustness

M

checks for the results.

Aggregate-level results show that FDI inflows complement public DI, whereas private DI is

ED

substituted by inward FDI. Conversely, outbound FDI has a positive effect on private DI,

PT

ranging in magnitude between 1.02 and 1.38. Sector-level analysis shows that FDI related to the oil sector contributes to public investments

CC E

in the GCC economies, but FDI into other sectors, such as manufacturing and services, has insignificant and unclear effects on DI. These results provide empirical evidence that FDI

A

inflow to GCC economies follows the production structure (oil-dependent) rather than leads it. The findings of this paper have implications for resource-rich economies in general, and GCC countries (oil states) in particular. These countries should diversify FDI into different productive sectors, such as finance and tourism, improving their financial institutions and giving more space for the private sector to engage with foreign firms, rather than letting foreign

22

investors be manipulated by their state-owned firms, which would limit transformation of technology. GCC countries should adopt screening policies to guarantee that FDI does not displace domestic firms; MNCs should also transform advanced technologies or introduce new products

IP T

to the country’s export basket. This process requires administrative skill to implement effective screening policies. Alternatively, these countries might adopt a fairly liberal system and then

U

SC R

pursue specific companies that fit well with the process of climbing the quality ladder.

N

References

A

1) Agosin, M. R. & Machado, R. (2005) Foreign investment in developing countries: does it crowd in domestic investment? Oxford Development Studies, 33(2), 149-162.

M

2) Aitken, B. J. & Harrison, A. E. (1999) Do domestic firms benefit from direct foreign investment? Evidence from Venezuela. American Economic Review, 605-618.

ED

3) Al-sadiq, A. J. (2013) Outward Foreign Direct Investment and Domestic Investment: The Case of Developing Countries. IMF working Paper, No.13/52 4) Al-Iriani, M.A., 2006. Energy–GDP relationship revisited : an example from GCC countries using panel causality. Energy Policy 34 (17), 3342–3350.

PT

5) Andersen, P. S. & Hainaut, P. (1998) Foreign direct investment and employment in the industrial countries.Bank for International Settlements, Monetary and Economic Department.

CC E

6) Arellano, M. & Bond, S. (1991) Some tests of specification for panel data: Monte Carlo evidence and an application to employment equations. The Review of Economic Studies, 58(2), 277-297. 7) Ayyagari, M. & Kosová, R. (2010) Does FDI facilitate domestic entry? Evidence from the Czech Republic. Review of International Economics, 18(1), 14-29.

A

8) Blundell, R. & Bond, S. (1998) Initial conditions and moment restrictions in dynamic panel data models. Journal of Econometrics, 87(1), 115-143. 9) Borensztein, E., De Gregorio, J. & Lee, J.-W. (1998) How does foreign direct investment affect economic growth? Journal of International Economics, 45(1), 115-135. 10) Cameron, A. C. (2010) Microeconometrics using stata, 2.Stata Press College Station, TX. 11) Chuang, W. B. (2007) FDI and domestic investment in Taiwan: An endogenous switching model. The Developing Economies, 45(4), 465-490.

23

12) Canton, E. and I. Solera (2016). Greenfield Foreign Direct Investment and Structural Reforms in Europe: What Factors Determine Investments? Luxembourg, European Economy Discussion Papers. 033. 13) Desai, M. A., Frtiz- FOLEY, C. & Hines, J. R. (2005) Foreign direct investment and the domestic capital stock. American Economic Review, 95(2), 33-38. 14) Doytch, N. (2016) Which sectoral FDI flows crowd in domestic entrepreneurship? Global Business and Economics Review, 18(2), 124-135.

IP T

15) Elbadawi, I. & Selim, H. (2016) Understanding and Avoiding the Oil Curse in Resource-rich Arab Economies. Cambridge: Cambridge University Press. 16) Farla, K., De Crombrugghe, D. & Verspagen, B. (2016) Institutions, foreign direct investment, and domestic investment: crowding out or crowding in? World Development, 88, 1-9.

SC R

17) Feldstein, M. & Horioka, C. (1980) Domestic Saving and International Capital Flows. The Economic Journal, 90(358), 314-329. 18) Feldstein, M. S. (1995) The effects of outbound foreign direct investment on the domestic capital stock, The effects of taxation on multinational corporationsUniversity of Chicago Press, 43-66.

U

19) Financial Times, (2016). fDi Markets database. FDI Intelligence, London.

N

20) Ghassan, H. B., & Alhajhoj, H. R. (2015). Long run relationship between IFDI and domestic investment in GCC countries. Journal of Economics and Econometrics, 59(2), 16–43.

M

A

21) Haddad, M. & Harrison, A. (1993) Are there positive spillovers from direct foreign investment?: Evidence from panel data for Morocco. Journal of development economics, 42(1), 51-74. 22) Hadri, K., 2000. Testing for stationarity in heterogeneous panel data. The Econometrics Journal, 3(2), pp.148-161.

ED

23) Hejazi, W. & Pauly, P. (2003) Motivations for FDI and domestic capital formation. Journal of International Business Studies, 34(3), 282-289.

PT

24) Hering, L., Inui, T. & Poncet, S. (2010) The Elusive Impact of Investing Abroad for Japanese Parent Firms: Can Disaggregation According to FDI Motives Help? Document de travail du CEPII(2010-01).

CC E

25) Herzer, D. (2008) The causal relationship between domestic and outward foreign investment: evidence for Italy. Applied Financial Economics Letters, 4(5), 307-310. 26) Herzer, D. & Schrooten, M. (2008) Outward FDI and domestic investment in two industrialized countries. Economics letters, 99(1), 139-143. 27) Hertog, S. (2013) The Private Sector and Reform in the Gulf Cooperation Council. London: LSE.

A

28) Hertog, S., 2012. National employment, migration and education in the GCC (Vol. 4). Gerlach Press. 29) Hanousek, J., Kocenda, E. and Vozarova, P., (2017). FDI's Impact on Inter-industry Interactions among Domestic Suppliers of Intermediate Goods. POLITICKA EKONOMIE, 65(4), pp.391409.

30) Morrissey, O. & Udomkerdmongkol, M. (2012) Governance, private investment and foreign direct investment in developing countries. World development, 40(3), 437-445.

24

31) Ni, B., Spatareanu, M., Manole, V., Otsuki, T. and Yamada, H., (2017). The origin of FDI and domestic firms’ productivity—Evidence from Vietnam. Journal of Asian Economics, 52, pp.5676.

32) Pedroni, P., 2004. Panel cointegration: asymptotic and finite sample properties of pooled time series tests with an application to the PPP hypothesis. Econometric theory, 20(3), pp.597-625. 33) Roodman, D. (2009) How to do xtabond2: An introduction to difference and system GMM in Stata. The Stata Journal, 9(1), 86-136. 34) Sahoo, A.K., Sahu, N.C., Sahoo, D. and Pradhan, B.B., 2014. Mineral export and economic growth in India: evidence from VAR model analysis. Mineral Economics, 27(1), pp.51-58.

IP T

35) Smarzynska Javorcik, B. (2004) Does foreign direct investment increase the productivity of domestic firms? In search of spillovers through backward linkages. The American Economic Review, 94(3), 605-627.

SC R

36) Stevens, G. V. & Lipsey, R. E. (1992) Interactions between domestic and foreign investment. Journal of international money and Finance, 11(1), 40-62.

37) Teece, D. (1977) Technology transfer by multinational firms: The resource cost of international technology transfer. Economic Journal, 87(346), 242-261.

U

38) Toone, J. E. (2012) Mirage in the Gulf: Examining the upsurge in FDI in the GCC and its legal and economic implications for the MENA region. Emory Int'l L. Rev., 26, 677.

A

N

39) Toews, G. and Vézina, P.L., 2017. Resource discoveries and FDI bonanzas: An illustration from Mozambique. Oxford Centre for the Analysis of Resource Rich Economies , working paper No.199

A

CC E

PT

ED

M

40) United Nations Conference on Trade and Development (2018), UNCTADstat, Retrieved from http://unctadstat.unctad.org/wds/ReportFolders/reportFolders.aspx

25

IP T SC R U N A M ED

A

CC E

PT

Figure 1. Impulse response function of domestic investment to oil prices shocks

26

Response to Cholesky One S.D. (d.f. adjusted) Innovations ± 2 S.E. Response of GCFGDP to OILPRICES

4

4

3

3

2

2

1

1

0

0

-1

-1 2

3

4

5

6

7

8

1

10

9

12

8

8

4

4

0

0

-4

-4

4

5

6

7

8

6

7

8

9

10

1

10

9

3

2

4

5

6

7

8

9

10

9

10

9

10

M

A

N

3

5

SC R

12

2

4

Response of OILPRICES to OILPRICES

Response of OILPRICES to GCFGDP

1

3

2

U

1

IP T

Response of GCFGDP to GCFGDP

Response to Cholesky One S.D. (d.f. adjusted) Innovations ± 2 S.E.

ED

Response of OILPRICES to OILPRICES

12 8

12 8 4

PT

4 0

CC E

-4 1

2

3

4

5

6

7

Response of OILPRICES to CREDIT

0 -4 8

9

10

1

2

Response of CREDIT to OILPRICES

6

6

A

8

4

4

2

2

0

0

-2

-2 2

3

4

5

6

7

8

4

5

6

7

8

Response of CREDIT to CREDIT

8

1

3

9

10

1

27

2

3

4

5

6

7

8

Figure 2. Impulse response function of domestic investment to oil prices shocks Response to Cholesky One S.D. (d.f. adjusted) Innovations ± 2 S.E. Response of OILPRICES to OILPRICES

10

5

5

0

0

-5

-5

1

2

3

4

5

6

7

8

9

10

1

Response of FDIGOILGDP to OILPRICES

2

SC R

10

IP T

Response of OILPRICES to FDIGOILGDP

3

4

5

6

7

8

9

10

4

3

3

2

2

N

4

U

Response of FDIGOILGDP to FDIGOILGDP

1

A

1

0 -1 2

3

4

5

6

7

8

9

10

A

CC E

PT

ED

1

M

0

28

-1

1

2

3

4

5

6

7

8

9

10

Response to Cholesky One S.D. (d.f. adjusted) Innovations ± 2 S.E. Response of OILPRICES to OILPRICES

Response of OILPRICES to FDIGNOILGDP

12

12

8

8

4

4

0

0 2

3

4

5

6

7

8

9

10

1

3

2

2

1

1

0

0

-1

-1 3

4

5

6

7

8

9

10

1

2

3

A

CC E

PT

ED

M

A

N

U

2

4

5

6

7

8

9

10

Response of FDIGNOILGDP to FDIGNOILGDP

3

1

3

IP T

Response of FDIGNOILGDP to OILPRICES

2

SC R

1

29

4

5

6

7

8

9

10

Table1- A. Definition and sources of data VARIABLES Outflows FDI/GDP ratio (stock) Inflows FDI/GDP ratio (stock)

Definition

Source of Data and definition

FDI stock is the value of the share of capital and reserves (including retained profits) attributable to the parent enterprise, plus the net indebtedness of affiliates to the parent enterprises.It is approximated by the accumulated value of past FDI flows.

UNCTAD

Saving/GDP ratio

World Bank Gross savings are calculated as gross national income less total consumption, plus net transfers. Gross fixed capital formation (formerly gross domestic fixed investment) includes land improvements (fences, ditches, drains, and so on); plant, machinery, and equipment purchases; and the construction of roads, railways, and the like, including schools, offices, hospitals, private residential dwellings, and commercial and industrial buildings. According to the 1993 SNA, net acquisitions of valuables are also considered capital formation. Annual percentage growth rate of GDP at market prices based on constant local currency. Aggregates are based on constant 2010 U.S. dollars. GDP is the sum of gross value added by all resident producers in the economy plus any product taxes and minus any subsidies not included in the value of the products. It is calculated without making deductions for depreciation of fabricated assets or depletion and degradation of natural resources. Inflation, as measured by the consumer price index, reflects the annual percentage change in the cost to the average consumer of acquiring a basket of goods and services that may be fixed or changed at specified intervals, such as yearly. The Laspeyres formula is generally used.

World Bank

SC R

GDP growth rate

World Bank

IP T

Gross Fixed Capital Formation/GDP ratio

Inflation rate

Trade openness ratio

World Bank

World Bank

U

Trade is the sum of exports and imports of goods and services measured as a share of the gross domestic product.

the sum of currency outside banks; demand deposits other than those of the central government; the time, savings, and foreign currency deposits of resident sectors other than the central government; bank and traveller’s checks; and other securities such as certificates of deposit and commercial paper. financial resources provided to the private sector by financial corporations, such as through loans, purchases of nonequity securities, and trade credits and other accounts receivable that establish a claim for repayment. For some countries, these claims include credit to public enterprises. The financial corporations include monetary authorities and deposit money banks, as well as other financial corporations where data are available (including corporations that do not accept transferable deposits but do incur such liabilities as time and savings deposits). Examples of other financial corporations are finance and leasing companies, money lenders, insurance corporations, pension funds, and foreign exchange companies. This type of foreign direct investment (FDI) where a parent company builds its operations in a foreign country from the ground up. In addition to the construction of new production facilities, these projects can also include the building of new distribution hubs, offices and living quarters. To resource sector

World Bank

N

Money supply/GDP ratio

World Bank

A

Financial Times, fDi intelligence unit

Table1- B. Descriptive Statistics

CC E

PT

FDI Greenfield in oil sector/GDP ratio FDI Greenfield in non-oil sector/GDP

ED

M

A

Domestic credit to private sector/GDP ration

VARIABLES

N

Mean

Outflows FDI/GDP ratio (stock) Inflows FDI/GDP ratio (stock) Saving/GDP ratio Gross Fixed Capital Formation/GDP ratio

66 66 66 66

12.94 24.40 45.56 24.53

Standard Deviation 12.29 19.86 14.77 7.189

GDP growth rate Inflation rate Trade openness Money supply/GDP ratio Domestic credit to private sector/GDP ration FDI Greenfield in oil sector/GDP ratio FDI Greenfield in non-oil sector/GDP

66 66 66 66 66 66 66

6.492 8.419 107.3 55.85 48.13 2.909 1.639

5.654 11.27 25.64 13.47 14.10 4.471 2.92

Source: Author’s work base on STATA outcomes

30

Table2. Panel unit- root test Hadri test 1st difference -0.98 0.62 0.53 -1.86 -0.263

Private domestic investments ratio

7.82***

-1.86

2.13*** 1.71*** 2.07*** 10.214*** 5.72***

GDP growth Saving ration Inflation rate Trade Money supply

all panels are stationary. H1: Some panels contain unit root

A

CC E

PT

ED

M

A

N

U

Ho:

SC R

FDI outflows ratio FDI inflows ratio FDI oil sector ratio FDI non-oil sector ratio Public domestic investments ratio

Level 11.04*** 9.79*** 2.53*** 4.89*** 3.63**

IP T

variables

31

-0.8009 -0.142 -0.511 -0.265 -1.410

Table3. Co-integration test- Pedroni test Panel roh

Panel PP

Panel ADF

PublicInvestment= f(FDI outflows, FDI inflows, saving, inflation, trade, money supply) PrivateInvestment= f(FDI outflows, FDI inflows, saving, inflation, tade, money supply) PublicInvestment= f(FDI oil sector, FDI non-oil sector, saving, inflation, tade, money supply) PrivateInvestment= f(FDI oil sector, FDI non-oil sector, saving, inflation, tade, money supply)

3.79***

---

-5.13***

4.23***

-7.14***

-5.40***

4.460***

0.679

0.2017

4.44***

-1.44*

-2.66***

H1: All panel are Co-integrated.

A

CC E

PT

ED

M

A

N

U

SC R

H0: No co-integration.

IP T

Equation(s)

32

Table 4 FDI and Domestic investments in GCC countries: Aggregate level analysis

FE ( private DI)

0.00915 (0.0949) -0.369** (0.142) 0.316*** (0.104) -0.131** (0.0530) 0.0353 (0.0460) 0.196** (0.0960) 0.144 (0.123) 0.0876 (6.701) 66 0.355 6 89.68 (0.000) ---

-0.0920 (0.112) 0.605*** (0.167) 0.0130 (0.123) 0.0160 (0.0624) -0.0456 (0.0541) 0.814*** (0.113) 0.137 (0.145) -4.139 (7.888) 66 0.816 6 33.66 (0.000) ---

FDI outflows/GDP Saving / GDP Inflation rate Trade openness Money supply GDP growth

M

Constant

ED

Observations R-squared Number of N F stat

22.80 (0.000) ---

44.34 (0.000) ---

CC E

AR(1)

PT

Wald test Hasuman test

N

FDI inflows/GDP

(4)

IV-FE (public DI) 0.318*** (0.0586) -0.431*** (0.126) 0.347*** (0.0551) -0.189*** (0.0628) 0.106*** (0.0298) 0.122 (0.0780) 0.126 (0.130) 2.775 (5.580) 66 0.60 6 ---

IV-FE (Private DI)

IP T

FE (public DI)

A

VARIABLES

(3)

SC R

(2)

U

(1)

Dependent Variables: Domestic investments

AR(2)

86.07 (0.000)

-0.59 (0.55) -0.61 (0.54)

114.41 (0.000)

-2.18 (0.03) -0.58 (0.56)

A

Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

33

-0.358*** (0.0775) 0.624*** (0.166) 0.194*** (0.0729) -0.0924 (0.0830) 0.0578 (0.0394) 0.560*** (0.103) -0.155 (0.172) 21.92*** (7.378) 66 0.82 6 ---

Table 5 FDI – Domestic investments: Sector-level analysis

Trade openness Money supply GDP growth

66 0.756 6 23.43(0.000)

34.52(0.000) --

32.31(0.000) --

--

--

CC E

AR(2)

66 0.439 6 5.91(0.000)

PT

Observations R-squared Number of N F stat (P-value) Wald test Hausman test AR(1)

IP T

0.590*** (0.165) 0.0771 (0.0573) 0.212*** (0.0548) -0.242*** (0.0665) 0.0551* (0.0329) 0.225*** (0.0607) 0.249* (0.137) 19.89*** (4.620)

ED

Constant

-0.411** (0.204) -0.0526 (0.0591) 0.0806 (0.140) 0.0298 (0.0732) -0.0286 (0.0620) 1.073*** (0.121) -0.0872 (0.151) -13.04 (8.768)

A

-0.0676 (0.229) -0.0908 (0.0797) -0.0469 (0.0762) -0.0129 (0.0925) 0.0880* (0.0458) 0.746*** (0.0844) -0.401** (0.191) 2.473 (6.426)

66

66

6 --64.32(0.000)

6 --185.44(0.000)

-0.14 (0.88) -0.24 (0.812)

-0.45 (0.86) -0.40 (0.68)

Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1

34

(8) IV-FE (Private DI)

SC R

Inflation rate

0.600*** (0.140) -0.0106 (0.0406) 0.151 (0.0960) -0.0828 (0.0503) 2.93e-05 (0.0426) 0.168** (0.0831) 0.256** (0.104) 5.630 (6.027)

U

Saving /GDP

(7) IV-FE (Public DI)

N

Non-Oil FDI/GDP

(6) FE( Private DI)

A

VARIABLES oil FDI/GDP

(5) FE (Public DI)

M

Dependent Variables: Domestic investments

Table. 6 VAR lag selection criteria AIC 33.10422 30.29570 29.21286 28.65566 28.03876 -28.81635*

SC 33.32416 31.61530 31.63213 32.17459 32.65735 -23.09808*

HQ 33.18098 30.75628 30.05725 29.88386 29.65077 -26.82052*

A

CC E

PT

ED

M

A

N

U

SC R

IP T

Lag 0 1 2 3 4 5

35