**2. Model**

Developing economies receive FDIs from various regions but they gain less from it because of institutional and infrastructural challenges. An economy with an established institutions and proper infrastructural systems is able to coordinate effectively with the flow of FDI [15–17]. In such economies, FDIs are properly allocated without compromising activities of local investors. FDIs are directed into areas of the economy where local investors have limited capacities to operate, thus widening the economic scope [18, 19]. Investors' confidence is high in such economies because of low operational cost and high investment

A dysfunctional institution creates unhealthy competition between foreign and local investors [20]. Unhealthy competition freezes the activities of local investors causing them to exit the market (because of capital and skills disadvantage), creating a foreign-dominated market. An effective institution properly coordinates FDI inflows across all sectors. According to the World Investment Report (WIR) (2012), about 60% of FDI inflows in Africa go to capital intensive activities such as mining and oil and gas activities [21]. Capital intensive requires high capitalization which

A review of growth literature highlights some studies on FDI. They examined the key determinants of FDI at the national level. Similar to what we highlighted in the foregone paragraph, institutions and infrastructure are the main prerequisite for affective FDI programs. For instance, Adewuni [1] examined Nigeria-China economic cooperation. The findings revealed a less than expected growth between FDI and economic growth, citing institutions and infrastructural and human capital as the main challenges. Kamara [21] in broader studies examined several Sub-Sahara African (SSA) economies. Busse and Groizard [15] also examined a national economy. Despite finding a positive growth relationship between FDI and economic growth, the finding also cited low human capital and weak infrastructural systems as the main drawbacks. However, AbuAl-Foul [22] found mixed outcomes in a dicountry study between Morocco and Tunisia. The economy of Morocco experienced a resilient growth link between FDI and growth while the economy of Tunisia experienced otherwise. All the studies gave insights in understanding FDI-growth relation, particularly at the national level [22]. However, there remains a gap at the

This chapter is examining Africa regional economy, looking at the impact of key macroeconomic indicators particularly China's FDI on regional economic growth using at least two decades of data. The remaining macroeconomic indicators include export, import, unemployment, and trade openness. Furthermore, the chapter is examining the impact of World and US FDI inflow on African economic growth using Granger causality test and autoregressive distributed lag (ARDL) model. The ARDL model will help test the short- and long-run effects of FDI on economic growth. Granger causality technique will help examine the causal relationship between economic growth and all the macroeconomic indicators. Finally, the chapter will look at whether Okun's law exists between

The outcome of this chapter has a two-fold effect; (1) inform policy regulators

The rest of the chapter is organized as follows. Section two explains the methods (i.e. Autoregressive Distributed Lag (ARDL) and Granger Causality) Section three

Sub-Sahara Africa, (2) Policy regulators will be able to make effective allocation of FDI resources to areas of greater impact in the economy. The recommendation session will offer some practical guidelines or policies that will boost the benefits of

about the actual empirical behavior of China's FDI on economic growth in

the local investors have no capacity to operate.

regional level that needs to be filled.

unemployment and economic growth.

**30**

FDIs in creating jobs and reducing inequalities.

returns.

*Regional Development in Africa*

### **2.1 Bound testing technique**

The vector auto-regression (VAR) of order p, denoted VAR (p), is expressed as the following [23]:

$$\mathbf{Y\_t = a + \sum\_{i=1}^{\rho} \mathcal{Q}\_{\mathbf{i}} \mathbf{Y\_{t-i} + \sum\_{i=1}^{\rho} \theta\_1 \mathbf{X\_t} + \mathbf{e\_t}} \tag{1}$$

where *yt* is the dependent variable presented by economic growth (RGDP), *xt* is a vector matrix representing explanatory variables, i.e., trade openness (OPENN), China's FDI inflow to Africa (CFDIITA), US FDI inflow to Africa (USFDIITA), China's export by Africa (CEBA), China's import by Africa (CIBA), *t* is trend variable, and others. According to bound model, *yt* must be an *I*ð Þ1 variable, but the independent variable *xt* must be stationary at either *I*ð Þ 0 or *I*ð Þ1 *:*

The vector error correction model (VECM) is expressed as follows:

$$
\Delta \mathbf{Y}\_{\mathbf{t}} = \mathbf{a}\_{\mathbf{t}} + \delta\_{\mathbf{t}} + \lambda \mathbf{Y}\_{\mathbf{t}-1} + \sum\_{i=1}^{\rho} \mathcal{Q}\_{1} \Delta \mathbf{Y}\_{\mathbf{t}-i} + \sum\_{i=1}^{\rho - 1} \theta\_{1} \Delta \mathbf{X}\_{\mathbf{t}-1} + \mathbf{e}\_{\mathbf{t}} \tag{2}
$$

where Δ is the first-difference order and *λ* represents the long-run multiplier matrix as follows:

$$
\lambda = \begin{vmatrix}
\lambda\_{\rm yy} & \lambda\_{\rm yx} \\
\lambda\_{\rm xy} & \lambda\_{\rm xx}
\end{vmatrix} \tag{3}
$$

The diagonal elements of Eq. (4) are unrestricted, so the selected series can be either I(0) or I(1). If, then Y is I(1). In contrast, if, then Y is I(0).

Eq. 2 is expanded to include all the regressors for the study, as shown below for later bound testing after estimation.

$$\begin{aligned} \Delta \ln \text{RGDP}\_{t} &= \mathsf{a}\_{t} + \sum\_{i=1}^{p} \mathsf{o}\_{i} \mathsf{d} \ln \text{RGDP}\_{t-i} + \sum\_{i=0}^{p} \mathsf{o}\_{i} \mathsf{d} \ln \text{ OPENN}\_{t-i} + \sum\_{i=0}^{p} \mathsf{v}\_{i} \mathsf{d} \ln \text{SSE}\_{t-i} \\ &+ \sum\_{i=0}^{p} \mathsf{o}\_{i} \mathsf{d} \ln \text{CEMA}\_{t-i} + \sum\_{i=0}^{p} \mathsf{o}\_{i} \mathsf{d} \ln \text{CIMA}\_{t-i} + \sum\_{i=0}^{p} \mathsf{o}\_{i} \mathsf{d} \ln \text{AFDIOTTH}\_{t-i} \\ &+ \sum\_{i=0}^{p} \mathsf{o}\_{i} \mathsf{d} \ln \text{CFDITTA}\_{t-i} + \sum\_{i=0}^{p} \mathsf{o}\_{i} \mathsf{d} \ln \text{FDIITA}\_{t-i} \\ &+ \sum\_{i=0}^{p} \mathsf{e}\_{i} \mathsf{d} \ln \text{USFDIITA}\_{t-i} + \pi\_{1} \ln \text{RGDP}\_{t} + \pi\_{2} \ln \text{OPENNN}\_{t} \\ &+ \pi\_{3} \ln \text{SSE}\_{t} + \pi\_{4} \ln \text{CEMA}\_{t} + \pi\_{5} \ln \text{CIMA}\_{t} + \pi\_{6} \ln \text{AFDIOTITW}\_{t} \\ &+ \pi\_{7} \ln \text{CFDIITA}\_{t} + \pi\_{8} \ln \text{FDIITA}\_{t} + \pi\_{9} \ln \text{USFDIITA}\_{t} + \pi\_{9} \end{aligned}$$

(4)

where ∅i, θi, γ, ϑi, ϑi, ∄i, ∈i, ρ<sup>i</sup> , δi, and φ<sup>i</sup> are short-run coefficients for the model and π1, π2, π3, π4, π5, π6, π7,π8, π9, π<sup>10</sup> are long-run coefficients.

ln RGDPt <sup>¼</sup> <sup>a</sup> <sup>þ</sup><sup>X</sup>

*DOI: http://dx.doi.org/10.5772/intechopen.89444*

ln OPENNt <sup>¼</sup> <sup>a</sup> <sup>þ</sup><sup>X</sup>

lnlnSSEt <sup>¼</sup> <sup>a</sup> <sup>þ</sup> <sup>P</sup>

ln CEBAt <sup>¼</sup> <sup>a</sup> <sup>þ</sup> <sup>P</sup>

ln CIBAt <sup>¼</sup> <sup>a</sup> <sup>þ</sup> <sup>P</sup>

ln AFDIOTTWt <sup>¼</sup> <sup>a</sup> <sup>þ</sup> <sup>P</sup>

ln CFDIITAt <sup>¼</sup> <sup>a</sup> <sup>þ</sup> <sup>P</sup>

ln FDIITAt <sup>¼</sup> <sup>a</sup> <sup>þ</sup> <sup>P</sup>

ln USFDIITAt <sup>¼</sup> <sup>a</sup> <sup>þ</sup> <sup>P</sup>

**3. Materials and method**

**3.1 Data and analysis**

**33**

p

*China-Africa Investments and Economic Growth in Africa*

<sup>∅</sup>ilnRGDPt�<sup>i</sup> <sup>þ</sup><sup>X</sup>

<sup>∅</sup>ilnRGDPt�<sup>i</sup> <sup>þ</sup><sup>X</sup>

<sup>∅</sup>ilnRGDPt�<sup>i</sup> <sup>þ</sup> <sup>P</sup>

<sup>∅</sup>ilnRGDPt�<sup>i</sup> <sup>þ</sup> <sup>P</sup>

<sup>∅</sup>ilnRGDPt�<sup>i</sup> <sup>þ</sup> <sup>P</sup>

<sup>∅</sup>ilnRGDPt�<sup>i</sup> <sup>þ</sup> <sup>P</sup>

<sup>∅</sup>ilnRGDPt�<sup>i</sup> <sup>þ</sup> <sup>P</sup>

<sup>∅</sup>ilnRGDPt�<sup>i</sup> <sup>þ</sup> <sup>P</sup>

<sup>∅</sup>ilnRGDPt�<sup>i</sup> <sup>þ</sup> <sup>P</sup>

p

i¼0

p

i¼0

p

i¼0

p

i¼0

p

i¼0

p

i¼0

p

i¼0

p

i¼0

p

i¼0

The research considered point annual FDI data but not accumulated stock data. Two models were used for the estimation, GMM and Granger causality method.

As a policy-based paper, the purpose of this chapter is to find an empirical justification for what has become a popular dialog in the economic environment, "China sudden interest in Africa." Has China's increasing presence in Africa via bilateral trade and investment link during the last two decades impacted significantly on Africa's macroeconomic indicators such as GDP per capita, unemployment, and human capital development. Considering Okun's law, there a link between economic growth and unemployment in the region? These questions arise because of the growing domination of China's investment in Africa vice verse that of the United States. Is the supposedly China's economic motive plan more effective and receptive to African economies than the US in addition to political motive? The

However, this Chapter is looking at the impact of FDI on key macroeconomic indicators in Africa using over decades of time series data from 1990 to 2014. The series include China's FDI inflow to Africa, China's Export to Africa, China's import from Africa, Secondary School Enrollment (SSE) (a measure of human capital), openness index, US FDI Inflows to Africa (USFDITA), World FDI inflows to Africa (WFDIITA) and African Investment Outflows to the World (AIOTTW). Annual

θ<sup>i</sup> ln OPENNt�<sup>i</sup> þ ε<sup>t</sup>

θ<sup>i</sup> ln OPENNt�<sup>i</sup> þ ε<sup>t</sup>

γilnSSEt�<sup>i</sup> þ ε<sup>t</sup>

φilnCEBAt�<sup>i</sup> þ ε<sup>t</sup>

∝ilnCIBAt�<sup>i</sup> þ þε<sup>t</sup>

δilnAFDIOTTWt�<sup>i</sup> þ ε<sup>t</sup>

(8)

ϑilnCFDIITAt�<sup>i</sup> þ ε<sup>t</sup>

∄ilnFDIITAt�<sup>i</sup> þ ε<sup>t</sup>

∈ <sup>i</sup> ln USFDIITAt�<sup>i</sup> þ ε<sup>t</sup>

i¼1

p

i¼1

p

i¼1

p

i¼1

p

i¼1

p

i¼1

p

i¼1

p

i¼1

p

i¼1

latter will be addressed in detail in the next chapter.

Eq. (4) also can be viewed as an ARDL of order (p, q, r). Eq. (4) indicates that economic growth tends to be influenced and explained by its past values. The structural lags are established by using minimum Akaike's information criteria (AIC). After regression of Eq. (4), the Wald test (F-statistic) is used to test the long-run coefficient to check whether it is significant or not. According to Pesaran et al. [23], the null and alternative hypotheses can be read as follows:

H0 ¼ π<sup>1</sup> ¼ π<sup>2</sup> ¼ π<sup>3</sup> ¼ π<sup>4</sup> ……………… *:*π<sup>10</sup> ¼ 0 NO LR Association ð Þ (5)

$$\mathsf{H}\_{0} \neq \mathsf{\pi}\_{1} \neq \mathsf{\pi}\_{2} \neq \mathsf{\pi}\_{3} \neq \mathsf{\pi}\_{4} \dots \dots \dots \dots \dots \dots \dots \dots \pi\_{10} \neq \mathsf{0} \ (\mathsf{LRAS association}) \tag{6}$$

The computed F-statistic value will be evaluated with the critical values tabulated in Table CI (iii) of Pesaran et al. [23] paper. As explained in Table CI (iii), the lower bound critical values assume the explanatory variables are integrated of order 0, or I(0), while the upper bound critical values assume the explanatory variables are integrated of order one, or I(1). Therefore, if the computed F-statistic is smaller than the lower bound value, then the null hypothesis is not rejected, which implies that there is no long-run relationship between economic growth and its determinants. However, if the computed F-statistic is greater than the upper bound value, then there is a long-run relationship between economic growth and its determinants. But, if the computed F-statistic falls between the lower and upper bound values, then the results are inconclusive.

#### **2.2 Granger causality**

Granger causality analysis is an analytical tool for examining whether a one-time series can correctly predict the other 21 ½ �. It is built on the premise that the future cannot predict the past because time does not travel backward. Theoretically, lag term of the independent variable is introduced into the model to statistically improve its prediction on the dependent variable as shown below:

$$\begin{aligned} \mathbf{Y\_t} &= \sum\_{\mathbf{n}=1}^{\rho} \mathcal{Q}\_{\mathbf{n}} \mathbf{Y\_{t-\rho}} + \sum\_{\mathbf{n}=1}^{\rho} \delta\_{\mathbf{n}} \mathbf{X\_{t-\rho}} + \mathbf{e\_t} \\ \mathbf{X\_t} &= \sum\_{\mathbf{n}=1}^{\rho} \gamma\_{\mathbf{n}} \mathbf{X\_{t-\rho}} + \sum\_{\mathbf{n}=1}^{\rho} \varrho\_{\mathbf{n}} \mathbf{X\_{t-i}} + \mathbf{e\_t} \end{aligned} \tag{7}$$

where *Yt* and *Xt* represent the two time series at t. *X*ð Þ *<sup>t</sup>*�*<sup>p</sup>* and *Y*ð Þ *<sup>t</sup>*�*<sup>p</sup>* represent the time series at time t–p, and p represents the number of lagged time points (order). ∅<sup>n</sup> and γ<sup>n</sup> are signed path coefficients. δ<sup>n</sup> and φ<sup>n</sup> are autoregression coefficients, while εtand ϵ<sup>t</sup> are residuals.

Peculiar to this study, mathematically, we introduce the lag of each series such as China's FDI inflows to Africa (CFDIITA), China's export to Africa (CEBA), China's import to Africa (CIBA), US FDI inflows to Africa (USFDIITA), openness (OPEN), and secondary enrolment (SSE) into equations for better prediction. Our model is thus expressed as follows:

*China-Africa Investments and Economic Growth in Africa DOI: http://dx.doi.org/10.5772/intechopen.89444*

where ∅i, θi, γ, ϑi, ϑi, ∄i, ∈i, ρ<sup>i</sup>

*Regional Development in Africa*

as follows:

inconclusive.

below:

**32**

**2.2 Granger causality**

and π1, π2, π3, π4, π5, π6, π7,π8, π9, π<sup>10</sup> are long-run coefficients.

Eq. (4) also can be viewed as an ARDL of order (p, q, r). Eq. (4) indicates that economic growth tends to be influenced and explained by its past values. The

criteria (AIC). After regression of Eq. (4), the Wald test (F-statistic) is used to

According to Pesaran et al. [23], the null and alternative hypotheses can be read

The computed F-statistic value will be evaluated with the critical values tabulated in Table CI (iii) of Pesaran et al. [23] paper. As explained in Table CI (iii), the lower bound critical values assume the explanatory variables are integrated of order 0, or I(0), while the upper bound critical values assume the explanatory variables are integrated of order one, or I(1). Therefore, if the computed F-statistic is smaller than the lower bound value, then the null hypothesis is not rejected, which implies that there is no long-run relationship between economic growth and its determinants. However, if the computed F-statistic is greater than the upper bound value, then there is a long-run

relationship between economic growth and its determinants. But, if the computed F-statistic falls between the lower and upper bound values, then the results are

Granger causality analysis is an analytical tool for examining whether a one-time series can correctly predict the other 21 ½ �. It is built on the premise that the future cannot predict the past because time does not travel backward. Theoretically, lag term of the independent variable is introduced into the model to statistically improve its prediction on the dependent variable as shown

<sup>∅</sup>nYt�<sup>ρ</sup> <sup>þ</sup><sup>X</sup>

<sup>γ</sup>nXt�<sup>ρ</sup> <sup>þ</sup><sup>X</sup>

time series at time t–p, and p represents the number of lagged time points (order). ∅<sup>n</sup> and γ<sup>n</sup> are signed path coefficients. δ<sup>n</sup> and φ<sup>n</sup> are autoregression

ρ

n¼1

ρ

n¼1

where *Yt* and *Xt* represent the two time series at t. *X*ð Þ *<sup>t</sup>*�*<sup>p</sup>* and *Y*ð Þ *<sup>t</sup>*�*<sup>p</sup>* represent the

Peculiar to this study, mathematically, we introduce the lag of each series such as China's FDI inflows to Africa (CFDIITA), China's export to Africa (CEBA), China's import to Africa (CIBA), US FDI inflows to Africa (USFDIITA), openness (OPEN), and secondary enrolment (SSE) into equations for better prediction. Our

δnXt�<sup>ρ</sup> þ ε<sup>t</sup>

(7)

φnXt�<sup>i</sup> þ ϵ<sup>t</sup>

Yt <sup>¼</sup> <sup>X</sup> ρ

Xt <sup>¼</sup> <sup>X</sup> ρ

coefficients, while εtand ϵ<sup>t</sup> are residuals.

model is thus expressed as follows:

n¼1

n¼1

H0 ¼ π<sup>1</sup> ¼ π<sup>2</sup> ¼ π<sup>3</sup> ¼ π<sup>4</sup> ……………… *:*π<sup>10</sup> ¼ 0 NO LR Association ð Þ (5) H0 6¼ π<sup>1</sup> 6¼ π<sup>2</sup> 6¼ π<sup>3</sup> 6¼ π<sup>4</sup> ……………… *::*π<sup>10</sup> 6¼ 0 LR Association ð Þ (6)

structural lags are established by using minimum Akaike's information

test the long-run coefficient to check whether it is significant or not.

, δi, and φ<sup>i</sup> are short-run coefficients for the model

$$\begin{aligned} \ln \text{RGDP}\_{t} &= \text{a} + \sum\_{i=1}^{P} \mathcal{Q}\_{i} \text{lnRGDP}\_{t-i} + \sum\_{i=0}^{P} \mathcal{Q}\_{i} \ln \text{ OPENN}\_{t-i} + \epsilon\_{t} \\ \text{In } \text{OPENN}\_{t} &= \text{a} + \sum\_{i=1}^{P} \mathcal{Q}\_{i} \ln \text{RGDDP}\_{t-i} + \sum\_{i=0}^{P} \mathcal{Q}\_{i} \ln \text{ OPENN}\_{t-i} + \epsilon\_{t} \\ \text{In } \text{RGSE}\_{t} &= \text{a} + \sum\_{i=1}^{P} \mathcal{Q}\_{i} \ln \text{RGDDP}\_{t-i} + \sum\_{i=0}^{P} \eta\_{i} \ln \text{RGS}\_{t-i} + \epsilon\_{t} \\ \text{In } \text{CEBA}\_{t} &= \text{a} + \sum\_{i=1}^{P} \mathcal{Q}\_{i} \ln \text{RGDDP}\_{t-i} + \sum\_{i=0}^{P} \eta\_{i} \ln \text{CEBA}\_{t-i} + \epsilon\_{t} \\ \text{In } \text{CIBA}\_{t} &= \text{a} + \sum\_{i=1}^{P} \mathcal{Q}\_{i} \ln \text{RGDDP}\_{t-i} + \sum\_{i=0}^{P} \eta\_{i} \ln \text{ICRHA}\_{t-i} + + \epsilon\_{t} \\ \text{In } \text{AFDIITTA}\_{t} &= \text{a} + \sum\_{i=1}^{P} \mathcal{Q}\_{i} \ln \text{RGDDP}\_{t-i} + \sum\_{i=0}^{P} \mathcal{Q}\_{i} \ln \text{AFDIITTA}\_{t-i} + \epsilon\_{t} \\ \text{In } \text{CFDIITTA}\_{t} &= \text{a} + \sum\_{$$

### **3. Materials and method**

The research considered point annual FDI data but not accumulated stock data. Two models were used for the estimation, GMM and Granger causality method.

#### **3.1 Data and analysis**

As a policy-based paper, the purpose of this chapter is to find an empirical justification for what has become a popular dialog in the economic environment, "China sudden interest in Africa." Has China's increasing presence in Africa via bilateral trade and investment link during the last two decades impacted significantly on Africa's macroeconomic indicators such as GDP per capita, unemployment, and human capital development. Considering Okun's law, there a link between economic growth and unemployment in the region? These questions arise because of the growing domination of China's investment in Africa vice verse that of the United States. Is the supposedly China's economic motive plan more effective and receptive to African economies than the US in addition to political motive? The latter will be addressed in detail in the next chapter.

However, this Chapter is looking at the impact of FDI on key macroeconomic indicators in Africa using over decades of time series data from 1990 to 2014. The series include China's FDI inflow to Africa, China's Export to Africa, China's import from Africa, Secondary School Enrollment (SSE) (a measure of human capital), openness index, US FDI Inflows to Africa (USFDITA), World FDI inflows to Africa (WFDIITA) and African Investment Outflows to the World (AIOTTW). Annual


*Descriptive*

 *statistics for all vectors.*

**DOPEN**

1

0.133681 0.117451

0.6689 0.4569

0.0583

0.0472

0.0762

1

0.4875

0.1330

1

0.2984

1

1

DOPEN

**35**

DSS DFDIITA

DCIDA DCFDIITA DAFDIOTTW

USFDIITA

UNEM

*The correlation matrix for all vector series from 1990 to 2014.*

*Source: author's own computation.*

**Table 2.** *Correlation*

 *matrix for each series from 1990 to 2014.*

0.4183 0.3837

0.2191

0.2461

0.1824

0.0444

0.0174

0.0339

0.15267

0.465755

0.417959

0.20233

0.113561

0.304829

0.201383

1

0.03539 0.0237

1

0.4628

1

**DSS**

**DFDIITA**

**DCIDA**

**DCFDIITA**

**DAFDIOTTW**

 **USFDIITA**

 **UNEM**

*China-Africa Investments and Economic Growth in Africa*

*DOI: http://dx.doi.org/10.5772/intechopen.89444*

**34**

#### *China-Africa Investments and Economic Growth in Africa DOI: http://dx.doi.org/10.5772/intechopen.89444*


**Table 2.** *Correlation matrix for each series from 1990 to 2014.*

**DAFDIOTW**

**34**

Mean Median

Max. Min.

Std. dev.

Skew.

Kurt.

J.B. Prob.

Obs. *Log of variables from 1990 to 2014.*

*Source: author's computation.*

**Table 1.** *Descriptive*

 *statistics for all vectors.*

22

21

22

22

24

25

25

24

25

21

0.756

0.805

0.868

 0.000

 0.379

0.560

0.434

0.282

 28.096

 1.939

2.423

3.699

3.100

 7.584

0.264

0.041

0.273

 1.552

0.730

0.810

0.131

 0.587

1.423

1.338

0.248

0.737

0.303

0.251 0.691

3.176

3.109 0.218 0.897

0.167

 0.812

 0.000

3.578

 0.416

 283.367

 1.605 0.448

3.987

 3.465

 17.661

0.227

0.807

 0.223

0.019

0.042

 0.018

 0.429 3.778

0.615

2.433

0.003

0.108

 0.010

 15.183

 19.918 0.929

1.352

2.220

0.303

 2.228

0.345

0.295

0.009

 0.252

0.202

0.322

0.034

 0.216

 **DCFDIITA**

 **DCEBA**

 **DCIBA**

 **DFDIITA**

0.122 0.075 0.727

0.075

0.082

 0.095

 17.457

 22.995

0.037

0.013

 0.049

 17.225

 22.069

0.037

0.010

 0.049

 17.109

 21.887

 **DGDP**

 **DOPENN**

 **DSSE**

 **UNEM**

 **USFDIITA**

*Regional Development in Africa*

#### **Figure 2.**

*Shows the unit root testing results for each of the series using the augmented Dickey-Fuller (ADF) technique. The series were stationary at different levels (either at I(0) or I(I)). The gaps represent missing data.*

FDI series (rather than FDI stock<sup>1</sup> ) is used in the analysis. Real GDP2 , served as the dependable variable, a measure of economic growth.<sup>3</sup> Trade openness stimulates economic growth.

**Table 1** shows the log description of the vector series. The average mean of real GDP is 0.037, less than the average mean of all regressors except China's export to Africa (CEBA) and openness (OPEN), which estimated average mean values of 0.034 and 0.010, respectively. The standard variance of all vectors is relatively a higher mean, suggesting high variation within vector indicators. Except for CIBA and UNEM, other vectors are not normally distributed (look at the Jarque-Bera test, they are not significant at 5%). In **Table 2**, there is a lower correlation between variables, suggesting a lower chance of perfect multicollinearity (**Figure 2**).

**Figure 1** shows the log stationary for each series for the Africa economy forms the period 1990 to 2010 employing the augmented Dickey-Fuller unit root test. We found a stationary path for all vector series at first difference I(1) except for UNEM and USFDIITA. The break within AFDIOTW, USFDIITA, and CFDIITA stationary paths was due to missing data series. CFDIITA, CIBA, USFDIITA, and CEBA also lost some data at the beginning.

### **4. Results analysis**

In this session, the analysis was based on the Pesaran [23] autoregressive distributed lag (ARDL) technique. There are two phases in the model: the long run and the short run [23]. Using real GDP (proxy by economic growth) as a dependent variable, the finding for both phases is presented in **Table 3**.

unemployment (UNEM) and US FDI inflow to Africa (USFDIITA) were stationary at the level I(0), while RGDP, China's export to Africa (CEBA), China's FDI to Africa (CFDIITA), openness (OPENN), secondary school enrollment (SSE), China's import from Africa (CIBA), Africa FDI outflow to the World (AFDIOTTW), and FDI inflow to Africa (FDIITA) were stationary at I(1). Furthermore, all the models satisfied the conditions for multicollinearity, i.e., there is no serial correlation. In Pesaran (2001), there are predefined critical values for making a statistical decision for the ARDL short- and long-run parameters [25]. There are different critical values for different significant levels depending on the structure of the model. Essential to this analysis is the critical values highlighted in Pesaran (2001), Table CII (iii) of page 303. It has an unrestricted intercept and no trend condition. At a 5% critical value, using ten (10) parameters point, Pesaran (2001) predefined the critical values in Table CII (iii). According to **Table 3**, critical values of 2.86 and 5.03 are decision results for an absolute lower I(0) and upper bound value I(1).

**Variables Short-run coefficients Variables Long-run coefficients**

D(RGDP(1)) 0.989 (0.232) D(RGD) 0.792 (0.193)

D(CEBA(1)) 8.513 (3.502) D(CEBA) 49.783 (13.383)

D(CFDITA(1)) 4.518 (0.712)\* D(CFDITA) 8.865 (1.579)\*

D(OPEN(1)) 8.908 (8.597) D(OPEN) 8.975 (8.500)

D(UNEMP(1)) 59.566 (12.056) D(UNEMP) 942.98 (182.909)\*

D(USFDIITA(1)) 0.658 (0.518) D(USFDIITA) 0.026 (0.770)

D(SSE(1)) 1026.498 (420.350) D(SSE) 170.536 (363.717)\*

D(FDIITA(1)) 1.783 (0.3554)\* D(FDIITA) 0.0124 (0.660)

D(CIBA(1)) 0.687 (2.641) D(CIBA) 13.413 (3.304)

D(AFDIOTTW(1)) 0.654 (1.272) D(AFDIOTTW) 5.798 (2.563)

R-squared 0.896 R-squared OBS: 23 F-computed 2.959 F-computed:18.94

*Absolute lower I(0) and upper bound value I(1) of 2.86 and 5.03. The standard errors are reported in brackets.*

Constant term 1100 (6.080)

*DOI: http://dx.doi.org/10.5772/intechopen.89444*

*China-Africa Investments and Economic Growth in Africa*

*Real GDP as dependent variable (Model 1).*

*Source: author's computation.*

*Dynamic ARDL Model Result.*

**Table 3.**

**37**

*Significance levels: \*\*\* if p < 1%, \*\* if p < 5%. and \* if p < 10% [22].*

Given the principle of ARDL model, each series must be either be stationary at first different I(1) or at the level I(0). According to Dickey-Fuller unit root test in **Figure 1**, all the series satisfied the ARDL condition, i.e., I(0) or I(1). For instance,

<sup>1</sup> It is broader and includes previous reserves and capital invested.

<sup>2</sup> GDP is a domestic-based indicator that measures the monetary value of all the finished goods and services produced within a country's borders in a specific time.

<sup>3</sup> Some use net export.


#### *China-Africa Investments and Economic Growth in Africa DOI: http://dx.doi.org/10.5772/intechopen.89444*

*Real GDP as dependent variable (Model 1).*

*Absolute lower I(0) and upper bound value I(1) of 2.86 and 5.03. The standard errors are reported in brackets. Significance levels: \*\*\* if p < 1%, \*\* if p < 5%. and \* if p < 10% [22]. Source: author's computation.*

#### **Table 3.**

FDI series (rather than FDI stock<sup>1</sup>

*Regional Development in Africa*

lost some data at the beginning.

**4. Results analysis**

<sup>3</sup> Some use net export.

**36**

economic growth.

**Figure 2.**

) is used in the analysis. Real GDP2

dependable variable, a measure of economic growth.<sup>3</sup> Trade openness stimulates

*series were stationary at different levels (either at I(0) or I(I)). The gaps represent missing data.*

*Shows the unit root testing results for each of the series using the augmented Dickey-Fuller (ADF) technique. The*

**Table 1** shows the log description of the vector series. The average mean of real GDP is 0.037, less than the average mean of all regressors except China's export to Africa (CEBA) and openness (OPEN), which estimated average mean values of 0.034 and 0.010, respectively. The standard variance of all vectors is relatively a higher mean, suggesting high variation within vector indicators. Except for CIBA and UNEM, other vectors are not normally distributed (look at the Jarque-Bera test, they are not significant at 5%). In **Table 2**, there is a lower correlation between variables, suggesting a lower chance of perfect multicollinearity (**Figure 2**).

**Figure 1** shows the log stationary for each series for the Africa economy forms the period 1990 to 2010 employing the augmented Dickey-Fuller unit root test. We found a stationary path for all vector series at first difference I(1) except for UNEM and USFDIITA. The break within AFDIOTW, USFDIITA, and CFDIITA stationary paths was due to missing data series. CFDIITA, CIBA, USFDIITA, and CEBA also

In this session, the analysis was based on the Pesaran [23] autoregressive distributed lag (ARDL) technique. There are two phases in the model: the long run and the short run [23]. Using real GDP (proxy by economic growth) as a dependent

Given the principle of ARDL model, each series must be either be stationary at first different I(1) or at the level I(0). According to Dickey-Fuller unit root test in **Figure 1**, all the series satisfied the ARDL condition, i.e., I(0) or I(1). For instance,

<sup>2</sup> GDP is a domestic-based indicator that measures the monetary value of all the finished goods and

variable, the finding for both phases is presented in **Table 3**.

<sup>1</sup> It is broader and includes previous reserves and capital invested.

services produced within a country's borders in a specific time.

, served as the

*Dynamic ARDL Model Result.*

unemployment (UNEM) and US FDI inflow to Africa (USFDIITA) were stationary at the level I(0), while RGDP, China's export to Africa (CEBA), China's FDI to Africa (CFDIITA), openness (OPENN), secondary school enrollment (SSE), China's import from Africa (CIBA), Africa FDI outflow to the World (AFDIOTTW), and FDI inflow to Africa (FDIITA) were stationary at I(1). Furthermore, all the models satisfied the conditions for multicollinearity, i.e., there is no serial correlation.

In Pesaran (2001), there are predefined critical values for making a statistical decision for the ARDL short- and long-run parameters [25]. There are different critical values for different significant levels depending on the structure of the model. Essential to this analysis is the critical values highlighted in Pesaran (2001), Table CII (iii) of page 303. It has an unrestricted intercept and no trend condition. At a 5% critical value, using ten (10) parameters point, Pesaran (2001) predefined the critical values in Table CII (iii). According to **Table 3**, critical values of 2.86 and 5.03 are decision results for an absolute lower I(0) and upper bound value I(1).

In **Table 3**, the dynamic model presented a computed F-value of 2.9595 for the short run. Based on the decision rule, it falls on the indecisive region, suggesting that China-Africa economic cooperation over the decades (in the short run) has not shown a significant effect on economic growth, i.e., indeterminate. Therefore, a lot is expected to be done for Africa's economic growth to experience the significance of China's FDI to Africa, China's export and import to Africa, US FDI to Africa, and Africa openness policies in the short run.

However, some selective series reported a decisive outcome. For example, Fcomputed value (22.47) for China's FDI inflows to Africa (CFDIITA) and FDI to Africa (FDIITA) critical value fell above the upper bound limit (5.03), suggesting a significant effect on regional economic growth. This is consistent with the theory that states that FDI stimulates growth [22]. This analysis has shown the significance of China's FDI on African growth.

In **Table 3**, in the long run, the F-computed value (18.94) fell above the upper limit of Pesaran [23] critical value (5.03), (i.e., it is in the acceptance region). The result suggests that overall change in all the series has a significant effect on economic growth. The estimated F-value (14.677) of unemployment, secondary school enrollment, and China's FDI to Africa is above the upper limit of the critical value, hence has a significant effect on economic growth. On this account, a change in China's FDI to Africa (CFDIITA) in the long-run will generally boost regional economic growth.

An individual analysis of each series or indicator reported different outcomes. In **Table 3**, all the series except China's FDI Inflows to Africa (CFDIITA) and Unemployment (UNEMP) in the long-run experienced an insignificant relationship with regional economic growth (at 5%). The impact of UNEMP on economic growth was positive at 5%. Highlighting the significance of education in economic development. A change in the quality of human capital will further boost regional economic growth.

Concerning China's FDI to Africa (CFDITA) series, in the long-run, a change leads to a fall in economic growth. A similar outcome was estimated in the shortrun. These outcomes give policy regulators a fair notion about the behavior of Chinese investments in Africa. If the evidence is considered enough, African economies must question the credibility of the investment. A contrary outcome was assumed before the analysis because of Chinese investment domination vice versa other developed economies. On this caveat, policy regulators in the region must review Chinese economic potentials and effectively attract investments from areas they have a comparative advantage or greater efficiencies that will benefit African economies. When successfully achieved, African economies will gain favorably from Chinese growing investments in the region. In the immediate term, governments across Africa can push most of China's FDI into developing labor-intensive programs or value addition industrial activities.

Africa (USFDIITA), China's Export to Africa (CEBA), and China's Import to Africa (CIBA) reported an insignificant effect on economic growth. The effect of openness (OPENN) was also insignificant. The latter does not imply African economies are closed or are not fully integrated into the global economy but rather suggests that less is gained from external participation. The majority of the exporting activities are primary-based. Primary-based activities have lower economic returns compared to capital or processed activities. In the last decade, the net wealth of African global

**Static model 2**

(350.941)\*\*

(8.226)

(0.966)

(7.734)

815.034

0.948

(99.78)\*\*

(0.523)\*\*

2.591

(1.438)\*

The US FDI inflow to Africa was found to be insignificant on African economic growth. This is due to the US losing interest in the African economy. The US FDI stock is by far the highest in African compared to China. But lately, it is declining while that of China is rising because of increasing China's interest in Africa. It is argued that the model of operation in Africa by China businesses is very much liked by the national governments because it is mainly economic. The US model, on the

trade inclusiveness was less than 10%. This figure is small compared to the combined volume of China and the United States, about 18% of world trade. Meanwhile, Africa is a significant player in the commodity and resource activities (led by Equatorial Guinea, Nigeria, Liberia, Kenya, Botswana, and Tunisia; mining:

Variables OLS regression Constant term 22..900

*China-Africa Investments and Economic Growth in Africa*

*DOI: http://dx.doi.org/10.5772/intechopen.89444*

D(CEBA) 14.446

D(CFDITA) 0.262

D(OPEN) 9.240

D(UNEMP) 397.358

D(USFDIITA) 0.884

D(SSE) 175.769

D(FDIITA) 2.217

D(CIBA) 2.453

D(AFDIOTTW) 3.567

R-SQUARED: 23

*The standard errors reported in brackets. Significance levels: \*\*\* if p < 1%, \*\* if p < 5%, and \* if p < 10%.*

Ghana, South Africa, and others).

*Static Model-Real GDP as Dependent Variable.*

*Source: author's computation.*

**Table 4.**

**39**

In model 1, a change in unemployment leads to a decline in economic growth in the long run. On the contrary, economic growth widens the unemployment gap in the region. This shows the lack of consistency between the flow of FDI and how it is applied in reducing unemployment. Mis allocation of investment harms development and widens the gap in inequalities and unemployment. Unemployment is a leakage in development and can affect economic growth negatively if loosely handled. Currently, at least 20% of the regional population remains unemployed. This paints a glooming picture for the region's future development if practical steps are not proposed today to boost job creation.

In **Table 4**, the impact of World FDI inflow in Africa (FDIIITA) on economic growth was positive. African FDI outflow to the World (AFDIOTTW) was found to have a positive effect on economic growth. On the contrary, the US FDI Inflows to


*The standard errors reported in brackets. Significance levels: \*\*\* if p < 1%, \*\* if p < 5%, and \* if p < 10%. Source: author's computation.*

R-SQUARED: 23

(1.438)\*

#### **Table 4.**

In **Table 3**, the dynamic model presented a computed F-value of 2.9595 for the short run. Based on the decision rule, it falls on the indecisive region, suggesting that China-Africa economic cooperation over the decades (in the short run) has not shown a significant effect on economic growth, i.e., indeterminate. Therefore, a lot is expected to be done for Africa's economic growth to experience the significance of China's FDI to Africa, China's export and import to Africa, US FDI to Africa, and

However, some selective series reported a decisive outcome. For example, Fcomputed value (22.47) for China's FDI inflows to Africa (CFDIITA) and FDI to Africa (FDIITA) critical value fell above the upper bound limit (5.03), suggesting a significant effect on regional economic growth. This is consistent with the theory that states that FDI stimulates growth [22]. This analysis has shown the significance

In **Table 3**, in the long run, the F-computed value (18.94) fell above the upper limit of Pesaran [23] critical value (5.03), (i.e., it is in the acceptance region). The result suggests that overall change in all the series has a significant effect on economic growth. The estimated F-value (14.677) of unemployment, secondary school enrollment, and China's FDI to Africa is above the upper limit of the critical value, hence has a significant effect on economic growth. On this account, a change in China's FDI to Africa (CFDIITA) in the long-run will generally boost regional

An individual analysis of each series or indicator reported different outcomes. In **Table 3**, all the series except China's FDI Inflows to Africa (CFDIITA) and Unemployment (UNEMP) in the long-run experienced an insignificant relationship with regional economic growth (at 5%). The impact of UNEMP on economic growth was positive at 5%. Highlighting the significance of education in economic development. A change in the quality of human capital will further boost regional economic

Concerning China's FDI to Africa (CFDITA) series, in the long-run, a change leads to a fall in economic growth. A similar outcome was estimated in the shortrun. These outcomes give policy regulators a fair notion about the behavior of Chinese investments in Africa. If the evidence is considered enough, African economies must question the credibility of the investment. A contrary outcome was assumed before the analysis because of Chinese investment domination vice versa other developed economies. On this caveat, policy regulators in the region must review Chinese economic potentials and effectively attract investments from areas they have a comparative advantage or greater efficiencies that will benefit African economies. When successfully achieved, African economies will gain favorably from Chinese growing investments in the region. In the immediate term, governments across Africa can push most of China's FDI into developing labor-intensive

In model 1, a change in unemployment leads to a decline in economic growth in the long run. On the contrary, economic growth widens the unemployment gap in the region. This shows the lack of consistency between the flow of FDI and how it is applied in reducing unemployment. Mis allocation of investment harms development and widens the gap in inequalities and unemployment. Unemployment is a leakage in development and can affect economic growth negatively if loosely handled. Currently, at least 20% of the regional population remains unemployed. This paints a glooming picture for the region's future development if practical steps are

In **Table 4**, the impact of World FDI inflow in Africa (FDIIITA) on economic growth was positive. African FDI outflow to the World (AFDIOTTW) was found to have a positive effect on economic growth. On the contrary, the US FDI Inflows to

Africa openness policies in the short run.

programs or value addition industrial activities.

not proposed today to boost job creation.

of China's FDI on African growth.

*Regional Development in Africa*

economic growth.

growth.

**38**

*Static Model-Real GDP as Dependent Variable.*

Africa (USFDIITA), China's Export to Africa (CEBA), and China's Import to Africa (CIBA) reported an insignificant effect on economic growth. The effect of openness (OPENN) was also insignificant. The latter does not imply African economies are closed or are not fully integrated into the global economy but rather suggests that less is gained from external participation. The majority of the exporting activities are primary-based. Primary-based activities have lower economic returns compared to capital or processed activities. In the last decade, the net wealth of African global trade inclusiveness was less than 10%. This figure is small compared to the combined volume of China and the United States, about 18% of world trade. Meanwhile, Africa is a significant player in the commodity and resource activities (led by Equatorial Guinea, Nigeria, Liberia, Kenya, Botswana, and Tunisia; mining: Ghana, South Africa, and others).

The US FDI inflow to Africa was found to be insignificant on African economic growth. This is due to the US losing interest in the African economy. The US FDI stock is by far the highest in African compared to China. But lately, it is declining while that of China is rising because of increasing China's interest in Africa. It is argued that the model of operation in Africa by China businesses is very much liked by the national governments because it is mainly economic. The US model, on the

other hand, has a strong element of political interference which is not very much liked by the national governments. But research shows that a stable environment promotes an effective growth [24].

On the contrary, there was no causality link between unemployment and economic growth. There was also no granger causality link between secondary enrollment (i.e., human capital) and economic growth. In theory, unemployment is a leakage in development when found to directly influence economic growth. Unemployment imposes a heavy burden on development via social intervention programs. As a result of that, minimizing the unemployment gap is a major aim for every economy. An increase in secondary school enrollment closes the literacy gap

The FOCAC cooperation has benefited China's economies more than it did for African because of growth hindering factors. In the form of foreign direct and portfolio investment, Chinese activities have grown in the region and are seen everywhere. China knows exactly what it wants from Africa while Africa is yet to

Is African economic performance growing as a result of China's economic cooperation or is yet to happen? This Chapter examined this question from the preview of FDI and growth analysis using at least two decades of FDI data. The chapter also examined the effect of US and the World FDI on growth using Autoregressive Distributive Lags (ARDL) and Granger Causality models. According to the ARDL model, there was a positive growth relationship between China's FDI and African economic growth in the long term but not the short term. It was positive for the World FDI inflow to African. However, the effect of US FDI inflows to Africa was

Change in human capital positively influences regional economic growth. There was no evidence of Okums Law as economic growth increases with unemployment, suggesting a lack of growth in the job market. Activities prevailing activities in the government and non-government sectors are not enough to bridge the gap in unemployment. The impact of openness i.e. economic inclusiveness was unexpectedly negative with economic growth in all models. This does not suggest, the region

The African community will gain significantly from China's investment engagements if the following recommendations are factored in the region growth plans. In the African economy, resolving growth issues are necessary, if gaining the most from FDI is the ultimate objective. Lingering growth problems will continue to hinder effective investment allocations. Without specifically outlining the core issues (facing development in the region) and actually resolving them is a recipe for underdevelopment. For example, Oil-producing nations need to go beyond crude oil activities which has a lower market price to processing activities which has a higher market price. Continuing with temporarily fixed and front-loaded deals with China will not resolve the region's major problems. China in particular knows that it wants from Africa and as a result deals with African in that regard. In the same vein, African economies need to know what they want to influence investment programs with China. They will be able to attract investments that will resolve their growth issues other than going for anything at all which has a long-run effect of collapsing

wakeup. Africa is still assuming it will gain from China engagement.

and increases the quality of human capital.

*China-Africa Investments and Economic Growth in Africa*

*DOI: http://dx.doi.org/10.5772/intechopen.89444*

**5. Conclusions and policy implications**

**5.1 Conclusion**

insignificant.

is a closed economy.

**41**

**5.2 Policy recommendation**

### **4.1 Granger causality test**

Granger causality test is used for forecasting between two series in an analysis. When two series are co-integrated then, there is likelihood of causality in at least one of the directions [25]. For instance, FDI and GDP are co-related, which implies that a change in one can cause the other to change, vice versa. To observe empirically the causality test for all the series in this study, Granger causality technique was employed. The results are presented in **Table 5**. Conclusively, there are situations of a unidirectional effect for some of the series.

**Table 5** summarizes the results for each series. According to **Table 5**, there is a uni-directional causality link between China's Export by Africa (CEBA), openness (OPEN), Africa's FDI Outflow Around the World (AFDIOTW), China's Import from Africa (CIBA), and China's FDI inflow to Africa (CFDIITA) on economic growth (GDP). This suggests that a change in any of the determinants or indicators will influence regional economic growth (but the reserve is not certain). As a result of this, proper allocation of FDI inflow from various sources including those from China and the US will directly boost regional economic growth. This certainly shows that Chinese investment in Africa is adding more to development. The ARDL findings in **Table 3** confirm this situation.


*China-Africa Investments and Economic Growth in Africa DOI: http://dx.doi.org/10.5772/intechopen.89444*

On the contrary, there was no causality link between unemployment and economic growth. There was also no granger causality link between secondary enrollment (i.e., human capital) and economic growth. In theory, unemployment is a leakage in development when found to directly influence economic growth. Unemployment imposes a heavy burden on development via social intervention programs. As a result of that, minimizing the unemployment gap is a major aim for every economy. An increase in secondary school enrollment closes the literacy gap and increases the quality of human capital.
