TELTheoretical Economics Letters2162-2078Scientific Research Publishing10.4236/tel.2018.814196TEL-88056ArticlesBusiness&Economics Econometric Analysis of Foreign Direct Investment in the Zimbabwean Mining Sector 2005-2014 PlaxedesGochero1*Open University of Mauritius, Moka, Mauritius1810201808143157317711, June 201823, October 2018 26, October 2018© Copyright 2014 by authors and Scientific Research Publishing Inc. 2014This work is licensed under the Creative Commons Attribution International License (CC BY). http://creativecommons.org/licenses/by/4.0/

The paper analyses the determinants and the effects of foreign direct investment inflows (FDI) in the Zimbabwe Mining Sector (ZMS) in a specific study for 14 minerals from 2005-14 estimating a random effect model. Mineral specific variables examined include capacity utilisation, volume of manufacturing index, labour cost, sectorial contribution to Gross Domestic Product (GDP), political instability, mineral price and mineral output. FDI inflow in the ZMS can be explained by capacity utilisation, volume of manufacturing index, labour, sectorial contribution to GDP and political instability. No statistical evidence could be established to support mineral price and output as major determinants of FDI in the mining sector. All these variables confirmed with literature except for volume of manufacturing index. As a result, the government is recommended to put in place supportive policies that encourage investments and recapitalization in the mining sector so that local firms can effectively compete at both domestic and international investment markets.

Foreign Direct Investment Inflow Zimbabwean Mining Sector1. Introduction

Access to foreign capital and investment enables a country to invest in human and physical capital as well as to make full use of opportunities otherwise not available. Acquisition of capital and making profitable investments is important in the economic improvement of a country. Foreign direct investment is therefore a vital component of development and operation at an international scale. Unfortunately, its empirical determinants and effects are not very well understood. Foreign direct investment (FDI) is crucial for developing countries because of the long-term finance, technology, technical knowhow, managerial expertise and marketing experience that if offers. It brings to the host country economic development, and leads to increased employment (Kukaj and Ahmet [1] ). Chimuka [2] , argued that FDI is crucial in the mining sector for carrying out mineral exploration, extraction, processing and marketing because Zimbabwe lacks enough capital and technological resources to finance such capital intensive large-scale projects. However, negative effects may also flow from FDI; these include increased financial risks, foreign exchange rates crises, transfers of obsolete and dirt technologies. As a consequence one needs to understand not only determinants but also effects of FDI.

A panel data analysis for 14 minerals across a 10-year period (2005-2014) is estimated in a bid to provide answers to the raised questions. Section 2 provides review of the theoretical and empirical literature of this topic. Section 3 shows methodological estimation procedures. Section 4 presents empirical results, discussion and their economic interpretation. Lastly Section 5 gives some possible policy recommendation, as well as suggested areas for future research.

2. Literature Review

Various theories have been put forward on the motivations and determinants of FDI. These theories can be divided into macro level theories and micro level theories. However, there is no established theory on the interrelationship between FDI and the mining sector hence it is necessary to consider FDI theories that try to provide an insight into what causes FDI to flow into various investment destination and not just FDI development theories. The eclectic paradigm that is also known as the OLI-Model which is an extension of internalization theory by Dunning [3] will be considered in the theoretical section of the paper because it is more acceptable than the heavily criticised neoclassical condition of perfect competition. The theory argues that the structure of an organisation is not the only significant factor in inviting FDI, but rather three additional elements were added to the theory, and these are Ownership, Location and Internalization advantages. It is a combination of three approaches to FDI, (O-L-I), “O” Ownership advantages, “L” Location and “I” for Internalisation, Denisia [3] .

Macroeconomic environment, political stability, FDI incentives, cost of labour, legal framework, availability of resources, level of infrastructural development are some of the factors considered before decisions are made by multinational companies(henceforth MNCs) as far as FDI location is concerned.

The eclectic theory takes into account the significance of three variables namely, country-specific, company-specific and internalisation relating to trade and FDI. The country-specific variables are geographical environment, the political environment, the government’s regulatory framework, taxation and fiscal policy, production costs and transportation costs and the cultural environment [3] . The specific advantages of each country can be divided into three categories, economic, political and social. The economic benefits consist of quantitative and qualitative factors of production, costs of transport, telecommunications, market size and infrastructure of the host economy, Zimbabwe has a wide range of these services however not much FDI has been attracted yet.

Internalisation, being another characteristic of the eclectic paradigm offers a framework for assessing different ways in which the company will exploit its powers from the sale of goods and services to various agreements that might be signed between the companies. The higher the cross-border market internalisation benefits the more the firm will want to engage in foreign production rather than offering this right under license, franchise. Eclectic paradigm OLI shows that OLI parameters are different from company to company and depend on context and reflects the economic, political, social characteristics of the host country. Therefore the objectives and strategies of the firms, the magnitude and pattern of production will depend on the challenges and opportunities offered by different types of countries. The main challenge in Zimbabwe can be argued to have been political and partly economical; this fact might have made the economy a less safe FDI destination although it provides all other advantages posed by the theory.

This scenario encourages the current study to assess the determinants of FDI, macroeconomic policy, political stability, and cost of labor, legal framework guiding FDI, availability of resources, FDI incentives and level of infrastructural developed to find out how they affect or influence the flow of FDI in the Zimbabwe mining sector. The study will test the theories of FDI effect on economic growth which have remained ambiguous in previous studies with the aim of shedding more light on this controversial topic. The study seeks to provide answers to the question of whether the mining sector should continue seeking or abandon the search for FDI. Economic development in the mining sector contributes to the economic development of the whole country and therefore the current research could help clear up the gray areas raised by previous researchers.

Empirical studies on both macro and/or micro data are yet to reach consensus on both the determinants and effects of FDI on the performances of either a particular sector or economy in general. Few studies on Africa have attempted either to establish determinants of FDI on a particular sector of the economy. Most of the available literature is on FDI and economic growth at large.

There are differing findings about the actual impact of political stability on FDI inflows. Demirhan and Masca [4] , studied FDI inflows in developing countries between 2000 and 2004 they came up with the conclusion that political risk has an insignificant effect on FDI. The argument was that investors were concerned with returns rather than political risks. Opposite results are held by Luiz and Ruplal [5] , who argued that security of tenure; political stability and the availability of infrastructure were the most important factors influencing the internationalisation of South African mining firms. It is apparent from these studies that political stability increases the probability of attracting more FDI inflows in a nation’s particular sector thus political stability is expected to have a positive effect on FDI inflows. Contrary to recent studies Tsikata et al. [6] , could not confirm that political instability was a significant determinant of FDI, Anyanwu [7] [8] , Moyo [9] , Kariuki [10] , later confirmed it as a prime variable. Divergent results need to further study the Zimbabwean mining sector as the developing country argument is contrary to the single nation argument. The argument is not the effect but rather the significance of the factors in influencing FDI.

Severiano’s [11] , study, “The determinants of FDI in Portugal,” concluded that most foreign investing firms do not choose to invest in Portugal to take advantage of a cheap labour force. According to Severiano [11] , a raise in the real minimum wage was found to affect positively FDI inflows. This is very interesting because previous research by different scholars indicated that FDI favours countries with low labour costs, see for example, Mody and Srinivasan [12] . Recent research by Bayraktar-Saglam and Böke [13] , however, examined the endogenous interaction between labour costs and Foreign Direct Investment (FDI) in the OECD countries via the Panel VAR approach under system GMM estimates for the period 1995-2009. The empirical findings revealed that sactorial composition of FDI and the decomposition of labour costs play a significant role in investigating the dynamic association between labour costs and FDI. It will be interesting to find out how labour cost determines FDI inflows to Zimbabwe.

Having discussed much about the determinants of FDI it becomes imperative to review the effect of FDI. Rate of growth can be explained by the state of technology used by a country [14] . However, on the state of technology it may be necessary to concentrate on sectorial level rather than at national level so as to have a clear understanding of its effects. Transfer of technology and knowledge by MNCs through FDI increases and improves productivity which leads to growth of Gross Domestic Product argue Varamini, and Vu [15] . Contrary to this view, MNCs could bring obsolete and dirty technology [16] .

Aswal [17] argues that host countries are integrated into the global economy thanks to financial flows received from FDI and that there is a link between the increase of FDI and rapid integration into global trade. However, further integration into the global economy can bring negative effects. Since FDI has a greater impact on imports than on exports this influences negatively the balance of payment which might prompt capital flight and repel FDI in the long run. FDI is the easiest way of spreading economic problems in the world especially those that have occurred in the MNCs countries of origin; for example, -host countries are forced to open economies and are more subject to global changes than would be the case, Khurtishi-Kastrati [18] .

Furthermore, FDI can be cardinal in producing a better economic environment resulting in economic growth: as Honsen and Rand [19] , argued FDI is a source of change. FDI can create entry barriers and or it can eliminate or reduce monopolies in these sectors changing the structure of the local economy for the best or worst. By takeovers and privatization MNCs force the adoption of their policies and procedures. These arguments have not only highlighted that the effects of FDI on the sector are inconclusive but even that the effects on the structure of the economy are not conclusive.

Using the OLS method of estimation Adewuni [20] , Anyanwu [7] , and Moyo, [9] , studied the possible impact of FDI on economic growth. Of these studies, Moyo [9] , made an attempt to establish the determinants of FDI in the post dollarization period of 2009 to 2012 in Zimbabwe using monthly data. The study established a positive relationship in respect to FDI. It should also be noted that Moyo [9] , study was based only on macro data. That data was based on the whole nation and not on sectors, which is instead the scope of this study. Adewuni [20] , employing a pooled OLS had a similar objective for nine African countries and concluded that economic growth was insignificant as a determinant of FDI. Anyanwu [7] , focused on the Nigerian economy for a 27-year period [1980-2007], and share the same results with Adewuni [20] , in the sense that economic growth was irrelevant in attracting FDI but rather he found natural resource endowment and exploitation especially for oil as the prime determinants of FDI. Since Nigeria is highly endowed with oil which was found to be a determinant for FDI, this study seeks to find out if such a conclusion could be reached on the diamond sector and FDI in Zimbabwe given the fact that the country is highly endowed with the germ.

Oluwatosin [21] , examined the causal linkages between FDI and economic growth and financial development in 5 African nations that is Gambia, Ghana, Ivory Coast, Nigeria and Sierra Leone from 1970-2005. A Vector Error Correction Model (VECM) was employed in the trivariate framework and results showed FDI to have bi directional causality on both financial and economic growth while financial development had a causal relation on economic growth and not the other way. A similar estimation concept was employed by Javorvick [22] , on firm-based panel of 9 MNCs. The focus was to test for productivity spillovers from FDI that would result from linkages between foreign and domestic suppliers and forward linkages (interactions between foreign suppliers of intermediate inputs and their domestic customers). This sort of methodology can also be employed to establish the relationship between mining sector performance and FDI. Given the size of the sector with 14 main minerals, a mineral based panel may also yield superior results to secondary macro data.

In a separate all Africa’s FDI determinant study done by Anyanwu [8] , for which data from 1996-2008 was studied for those nations in a pooled ordinary squares and Fixed General Least Square estimation technique was employed. Results showed market size and trade openness to be the most (positive) significant factor in attracting FDI. However financial market development was found to have a negative impact while economic growth had no share on attracting FDI.

In a study by Kariuki [10] , an estimated Least Square Dummy Variable also known as the fixed effects model for a sample of 35 African countries including Zimbabwe was used and produced similar results to those found by Anyanwu [8] . The only difference was that financial sector development this time around was found to have a positive effect on FDI attraction for the nation. One would wonder if the use of a different methodology would influence the results or if it is the sample size that affects variable significance. It is not only Anyanwu [8] , who could not establish the significance of the economic growth similar results were obtained even by Adewuni [20] .

In terms of determinants and effects of FDI on mining sector nothing much can be drawn from the reviewed literature as most studies focused on either or both determinants and effect of FDI on economic growth. Interestingly, of that existing literature contradictory views on the link between FDI and economic growth were found. Hence, this study does not only seek to contribute to that debate but to empirically establish determinants and effects of FDI in the Zimbabwean mining sector context.

3. Methodology

Theory and empirical studies carried out by previous scholars provide a foundation for model building. Approach and methods to be used in this research are explained in this section. The approach and methods of this study relates to studies by Javorvick [22] and Oluwatosin [21] . This study presents results from panel data regressions from 14 minerals for the period 2005 to 2014. The suitable model to study determinants and effects is the one expressed by an analytical approach similar to that used by Javorvick [22] , Oluwatosin [21] , with the needed ad hoc adjustments model is:

F D I i , t = α + β X i , t + ε t (1)

X i , t = β 0 + D P S i , t + β 2 V M I i , t + β 3 C O S T i , t + β 4 P R I C E i , t + β 5 C U i , t + β 6 C O N T R i , t + β 7 O U T i , t + β 7 L A B R + μ i , t (2)

D = { 1   2005 - 8 ,   2013 0   otherwise

where FDI is the foreign direct investment inflow in a particular mineral, X is a set of all possible factors that determine FDI inflow into a firm and t is annual time period from 2005 to 2014 for 14 minerals. While α and β are related coefficients and ε t is a vector of white noise disturbance.

The study wants to establish through estimation the determinants and effects of FDI on the mining sector performance of the Zimbabwean economy.

3.1. Panel Data Analysis

Panel data analysis is a frequently used approach in FDI and growth research since it enables the researcher to study the dynamics of the change of economic growth per capita for a short time series [2005-2014]. Since panel data combines both cross-sections and time series it can enhance the quality of data and sort out economic effects that cannot be distinct with only cross-sections or time series data. Panel data also has more numbers of data points that generate additional degrees of freedom which improve the efficiency of the econometric estimates. Moreover, using information of both temporal (time) and minerals (cross-section) effects one can substantially overcome omitted or missing variables problem [23] . Some models that can be used for panel data analysis are described below.

3.2. Constant Coefficient Model

F D I j , t = a + β i X j , t + ε j , t ,

j = 1,∙∙∙14, j is for minerals; t = 1∙∙∙10, for 2005 = 1.

The constant intercept assumption means that all minerals are considered to be the same and there are no significant cross sectional and time specific effects. The first step here is simply combining both the time-series and cross-section data, also known as pooling, and then estimating parameters with ordinary least squares (OLS). This simple use of pooled data enables the testing of theories and assumptions with relative ease. The estimated parameters of this model will be used as a benchmark against other estimates from the more sophisticated models.

By ignoring the cross section and time specific effects that possibly exist, the parameters estimated will be meaningless and inconsistent. To account for possible heterogeneity among minerals, fixed effects models and random effects models are considered to be more appropriate for handling panel data.

Hsiao [23] argued that it makes no difference whether fixed or random effect models are used when T (time series) is large, but if T is finite and N is large it could make a surprisingly difference in the estimates. Since the data set of this study only consist of 10 years for each of the 14 minerals, it is essential to choose the correct model to make the best use of this small amount of information. Hsiao [23] argued that one way to decide to use fixed effects or random effects model is to test for misspecification of the random effects model, where a is assumed to be random and uncorrelated with the independent variables.

To test the significant hypothesis, this study will perform a Hausman-test which tests for correlated random effects. If a is uncorrelated with the independent variables and thus the null hypothesis holds then a random effects model should be applied. Contrary, if the alternative hypothesis holds, a fixed effect model should be applied. The Hausman-test suggests the null hypothesis may not be rejected if, P-value > critical value mostly 0.05 thus it will be possible to use cross-section random effects estimators.

3.3. Definition and Justification of Variables

The determinant variables of FDI inflows include macroeconomic policy, political stability, mineral price, cost of labour, mineral output, volume of manufacturing index and sectorial contribution to economic growth. These variables are interrelated and are derived from the eclectic theory by Dunning [3] .

Foreign direct investment inflow (FDI) constitutes a direct capital investment for the purpose of development and enterprising in a foreign country by TNCS. Such investment is very useful in developing countries especially when channeled to a nation’s industries such as mining. For enterprises on Zimbabwe Stock Exchange such activities are not considered FDI but rather transfer payment. FDI in the mining sector like Output in the very sector will be considered on annual bases and the variable is also considered to be a dependent variablehence no prior exception to talk about, Henry [24] , Javorvick [22] and Oluwatosin [21] .

Political stability (PS) is cardinal for normal macroeconomic and business environment in a nation [5] [25] . Political risk refers to activities that interrupt business or harm staff and property for example property take over by government and violent demonstrations. These factors influence the decision to invest or not to invest in a certain location Dunning [25] . All the above will be considered as proxy for political stability. Political stability increases the probability of attracting more FDI inflows into nation.

Cost of labour (COST) is the total amount of expenditures on labour force, cost of labour is considered as a percentage of income in that particular year. Henry [24] , Javorvick [22] , and Oluwatosin [21] argued that cost of labor theories are important for FDI as investors look for countries with lower labor costs so as to maximize profits. Thus we would expect cost of labour to negatively influence FDI inflows in a firm.

Output (OUT) is the real output of each mineral produced annually and this would be introduced to capture the total output of economic activities in the firm, This variable is also used as the income level, which is considered as the main resource of technological development and human capital improvement. Investors would also consider this variable especially as percentage of total sector output so as to measure the potential market size when deciding FDI location, especially those who target to enhance their market share in the host country. All the extracting mining firms in the economy are viewed to constitute Zimbabwe Mining Sector (ZMS).

Price-PRICE is the closing year International price of the mineral understudy. This variable is used to test the impact of mineral prices in a small economy such as Zimbabwe and it would be expected that this variable would influence the attraction of FDI. Adewuni [20] , and Anyanwu, [7] , have all included the mineral price in their study of determinants of FDI in a particular sector or for that particular resource.

Capacity utilisation-CU is the measure of operational capacity for minerals by the mining firms and it would be expected to impact positively on direction of foreign direct investment inflows. However, there might be a possibility of reserve effect which is beyond the scope of the study. The variable was once used by Ramirez [26] , Anyanwu [8] , and Fasanya [27] .

Volume of the Manufacturing Index-(VMI) is considered the proxy for rate of growth of mining sector as the index measure for mineral and quarry activities only. This variable provides an indication on the possible effect of the sectorial activities on its ability to influence the direction of flow of FDI in that particular sector. Hermes and Lensink [28] , Varamini, and Vu [15] , Volume of the Manufacturing Index-(VMI) is considered the proxy for rate of growth of mining sector as the index measure for mineral and quarry activities only. This variable provides an indication on the possible effect of the sectorial activities on its ability to influence the direction of flow of FDI in that particular sector. Hermes and Lensink [28] , Buckley, et al. [29] , Ikechi, et al. [30] , Orji, et al. [31] , are among those who have used this variable.

Contribution to GDP-(CONTR), Oladipo [32] , Fasanya [27] , and Anyanwu [8] , Moyo [9] all argued that contribution of a particular sector has a role to play in influencing the direction of the flow of FDI in that particular sector. They all agreed that the variable positively help to attract FDI hence; a positive coefficient on the variable is expected. This will be measured as the percentage contribution of the mining sector to the country’s GDP.

4. Estimated Results4.1. F-Test (Pool Ability Test)

In a bid to find the most suitable model between the fixed effects model and the pooled OLS model pool ability tests were carried out. The hypothesis to be tested is that there is no mineral specific heterogeneity i.e. the pooled OLS model is the most suitable against the alternative that the fixed effect is the proper model. The F-test results1 reject the null hypothesis, implying that there is mineral specific heterogeneity, thus the pooled OLS model is not the appropriate model of the two.

4.2. Breusch and Pagan Lagrange Multiplier Test for Random Effects (LM)

To determine the most suitable model between the random effects and the pooled OLS model a LM test was carried out. Results2 failed to accept the null hypothesis and it can be concluded that there are panel effects. Thus the pooled OLS model is not to be preferred over the random effects model.

4.3. Hausman Test

1Annual bases for each firm in the firm panel model and FDI is a depended variable.

2See appendix for the Breusch and Pagan Lagrange multiplier test for random effect results.

3See appendix for the Hausman test results.

The Hausman test3 was carried out to determine the best model between the fixed effects model and the random effects model. It was necessary to test hypothesis that there is no relationship between mineral specific heterogeneity and the explanatory variables i.e. the random effects model is the appropriate model against the alternative hypothesis that the fixed effects is the appropriate model. The null hypothesis could not be rejected and therefore the conclusion is that the random effects is the best model.

4.4. Heteroskedasticity Test

The Breusch-Pagan/Cook-Weisberg test for heteroskedasticity4 was conducted to verify the nature of the variances of the error terms, that is, are they homoscedastic or not? It was decided to reject the null hypothesis and conclude that there is statistical evidence of heteroskedasticity. However, as noted in Mulenga [33] , this problem can be easily dealt with by estimating a feasible generalized least square (FGLS) in the random effects model.

4.5. Estimation of the Model4.5.1. The Unrestricted Random Effects Model (REM)

The various econometric tests conducted concluded that the REM is the most appropriate model for estimating the mineral specific determinants of FDI in the mining sector. A Generalized Least Squares (GLS) regression of the unrestricted REM was run. The results found all other variables significant at the conventional levels of significance (1%, 5% and 10%) except for price and costs. It was therefore decided to drop these variables one by one, starting with price as it was the variable that was highly insignificant. Subsequently cost was dropped and another regression was run.

4.5.2. Presentation of the Restricted REM Results

Tables 1-3 are a summary of the REM results after elimination of insignificant variables of the unrestricted model. The explanatory variable of the restricted model are all significant at either the 1%, 5% or 10% level of significance.

Random-effects (GLS), using 140 observations;

Included 14 cross-sectional units;

Time-series length = 10;

Dependent variable: L_FDI;

“These cross-sectional units were the results of REM using Stata. There were 140 observations for the 14 minerals. The regression covered 10 years.”

There were 140 observations obtained from pooling time series data from 14 minerals each contributing 10 observations as reported in the table above. The regression results report a Wald chi 2(5) statistic of 21.03, with a p-value of 0.008. This implies that there is statistical evidence that at least one of the regression coefficients is statistically different from zero. The reported Hausman test further supports the argument early raised that random effect model is the best one and its estimates are consistent.

4.6. Capacity Utilisation (CU)

4See appendix.

The variable CU is found to be significant at 5% level with the expected positive sign. The results confirm the findings of many other studies done on the country specific determinants of foreign direct investment inflows. For instance, Ramirez [26] , Anyanwu [8] and Fasanya [27] all found a positive value for the CU variable with respect to FDI. A positive coefficient of 0.0224 implies that a 1%

REM results
CoefficientStd. Errort-ratiop-value
CONST8.598211.440975.9670<0.0001***
PS−1.022680.185225−5.5213<0.0001***
L_LABR0.4636140.2074142.23520.0271**
VMI−0.04078230.00778521−5.2384<0.0001***
CU0.02241090.009398142.38460.0185**
CONTR0.3474870.021728115.9925<0.0001***

***Significant at 1% level of significance. **Significant at 5% level of significance. *Significant at 10% level of significance. Source: Created by author.

Breusch-Pagan test
Null hypothesis: Variance of the unit-specific error = 0
Asymptotic test statistic: Chi-square (1) = 392.6 with p-value = 2.2478e−087

Source: Created by author.

Hausman test
Null hypothesis: GLS estimates are consistent
Asymptotic test statistic: Chi-square (5) = 0.71204 with p-value = 0.982301

Source: Created by author.

increase in the capacity utilisation in the mining sector will result in an increase in FDI inflows share by approximately 2.24% holding everything else constant. The results confirm that the economic size of the mining sector is very important for the attraction of FDI inflow in that particular sector. The larger the sector the bigger the markets power, this fosters innovativeness amongst investors as they try to secure more of the input so as to differentiate their final products from their competitors. According to Krugman [34] , through product differentiation, firms can concentrate on a limited set of products, this result in lower per unit cost as firms exploit economies of scale. Hence increase in capacity utilisation has positive effect in attracting FDI inflows in the Zimbabwe’s mining sector.

4.7. Contribution of the Mining Sector (CONTR)

The variable is highly significant at 1% level with the expected positive sign. The coefficient of 0.347 implies that for a 1% increase in the contribution of the mining sector to the economies GDP leads to 34.47 percentage increase in FDI inflows in the mining sector. The results confirm economic theory, as it is expected that the more the sector grows the more it can attract investments. Anyanwu [8] , Fasanya [27] and Oladipo [32] all get of similar results for different nations and sectors while they agree that contribution of sector influence FDI inflows in that particular sector.

4.8. Volume of Manufacturing Index (VMI)

The variable VMI is found to be highly significant at 1% with an unexpected negative sign. The negative coefficient of 0.04078 implies that a 1% increase in the VMI will result in a decrease in FDI in the mining sector share by approximately 4.08%, holding everything else constant. Orji, et al. [31] , results were that FDI has a negative correlation with manufacturing output. Ikechi, et al. [30] , study established that Nigerian industrial productivity is not dependent on FDI, however, when individual component sectors were examined the findings were that FDI has a positive and significant relationship with mining sector productivity in Nigeria at 5% level in the short term. The results, though not conforming to economic theory, could be justified on the grounds that Zimbabwe for the large part of the period understudy experienced economic decline. Thus it can be concluded that economic decline negatively affects FDI inflow in the ZMS.

4.9. Labour (LABR)

The variable was found to have the expected positive sign and statistically significant at 5% level. The coefficient of 0.4636 implies FDI inflows in the mining sector are relatively elastic to changes in labour level available to the sector and mineral specific. Thus a unit percentage increase in Labour will result in 46.36 percentage increase in FDI inflows in the ZMS. Thus a unit percentage increase in Labour will result in 46.36 percentage increase in FDI inflows in the ZMS. The findings are similar to those held by Henry [24] , Demirhan and Masca [4] , Hussain and Kimuli [35] . These findings lead to the conclusion that Labour levels availability has been positively affecting and influencing FDI inflow in the Zimbabwe Mining Sector for the period under study.

4.10. Political Stability (PS)

Dummy variable for political instability is highly significant at 1% level of significance. The variable has the expected negative sign and the coefficient of 1.02268 implies that political instability negatively affects the FDI inflows in the Zimbabwe’s mining sector by 63.037 percentage. Similar results are held by Demirhan and Masca [4] , and Luiz and Ruplal [5] for the different nations they studied.

5. Conclusions

Comparative results of the three models: the pooled OLS (POLS), random effects model (REM) and the fixed effects model (FEM), show that the REM and POLS report results which are comparably similar in terms of the coefficients and expected signs. However, CU was also significant but the coefficient of this variable for the FEM differs significantly from those of the POLS and the REM. Legislators have well vested interest in implementing policies aimed at stabilisation and efficiency in the Mining sector. For one to formulate an effective policy that strikes a balance between the two, one needs to understand the determinants of FDI so as to adopt policies that promote a sustainable and stable Mining sector for economic growth.

The study finds that mining firms can increase FDI inflows by a small percentage increase in capacity utilisation. Hence there is a call for mining firm managers to be innovative and incorporate information and communications technology when dealing with mineral products.

It is thus recommended that government should put in place supportive policies that encourage investments and recapitalization of the mining sector so that local commodity can effectively compete both on the domestic and international markets. This could involve provision of a relative stable political environment which boosts investor confidence and thus helps attract foreign direct investments. Furthermore, government should put in place an enabling environment with enough incentives to spur innovativeness amongst local mineral mining firms, so that different varieties of products appealing to different sections of consumers both domestically and internationally are readily available.

As is noted in the Chamber of Mines annual reports [36] , most companies are operating below full capacity owing to challenges related to working capital, inadequate provision of supportive infrastructure (energy, water), obsolete equipment and low consumer demand. The Zimtrade export capacity mining survey of 2011 noted a decline in companies exporting; thus there is need for government to pursue a rigorous export led industrialization strategy following in the footsteps of the Asian tigers. However, it is imperative to put place an enabling policy environment which recognizes need to tap in foreign direct investments, protect property rights and uproot corruption.

From the results, labour was found to be a significant explanatory variable for FDI inflow in the ZMS. It is thus recommended that government should undertake policies which nurture and support growth of labour to the sector. This would require the government to establish a menu of policies that reinforces confidence amongst domestic and foreign investors, spruces the image of Zimbabwe as a prudent borrower and above all, fosters the development and retention of human capital. Due to data limitations, the study concentrated on 14 out of 66 minerals to draw conclusions on the mineral specific determinants of FDI in the country’s mining sector. A richer data set would broaden this study to include other important minerals in the sector, such as diamond. Furthermore, the scope of the study was limited to investigating the sector mineral specific determinants of FDI and this could be extended to incorporate company specific determinants of FDI; for instance the analysis could be enriched adding the capital to labour ratios and proxies for productivity.

Conflicts of Interest

The author declares no conflicts of interest regarding the publication of this paper.

Cite this paper

Gochero, P. (2018) Econometric Analysis of Foreign Direct Investment in the Zimbabwean Mining Sector 2005-2014. Theoretical Economics Letters, 8, 3157-3177. https://doi.org/10.4236/tel.2018.814196

Appendix 1

The Sample

The sample was made up of the following minerals, asbestos, chrome, coal, cobalt, copper, gold, graphite, iridium, nickel, palladium, phosphate, platinum, rhodium, and rhithium. The original mineral data obtained from Zimbabwe National Statistical Agency, (ZIMSTAT) is attached.

Raw data
FDIOUTPRICEvmicucontrlabrPSCRS IDTIMEID
ASB 1133,823.7486554.601542.64769256.72665642154.993956012005
ASB 21,581,273.7126665.27197.68107849.1765461841.0648012006
ASB 3615,281.64629.034.72989926146.18964081815.0138012007
ASB 41,059,822.5563571.84173.6934.03323461845.8448012008
ASB 5793,713.2644970.81315.300311231.91864181973.4708112009
ASB 61,615,114.22030.97644.41685541.227151122246.1698112010
ASB 72,551,880.4360366.84665.302755132649.4584112011
ASB 85,952,849.480798.305084753.160152421.309112012
ASB 96,145,124.98377834.392882762.85395815.72450.228112013
ASB 106,152,81602.17554E−0963.25656016.92152.3862112014
CHR 1273,246.816544.3882142.072228656.72665644400.155022022005
CHR 23,228,709.504519.60487.2535298349.1765463759.1616022006
CHR 31,256,307.2398.3646084786.76832146.18964083705.9696022007
CHR 42,163,989.152253.9919734.03323463768.9216022008
CHR 51,620,636.288193,673.7494.8797114231.91864184029.5136122009
CHR 63,297,806.4516,776.08110.061926241.227151124586.3216122010
CHR 75,210,534.112599,079.38250.9965.302755135409.7728122011
CHR 812,154,772.16408,475.81119.879701153.160154943.928122012
CHR 912,547,368.16355,142102.543692462.85395815.75002.976122013
CHR 1012,563,072408,4224.7778E−1163.25656016.94394.8304122014
COAL 18398.988638.6525.3558375556.7266564135.2506667032005
COAL 299,243.1210,826.5591422.3327131549.176546115.548032006
COAL 338,61610,524.34548.404119746.1896408113.913032007
COAL 466,516.0613,065.58374.26881434.0332346115.848032008
COAL 549,814.6412,482.734134.7701760831.9186418123.858132009
COAL 6101,3672500.16936.3557694541.22715112140.973132010
COAL 7160,159.862562.054173.007356965.30275513166.284132011
COAL 8373,609.81593.61347.000786153.16015151.965132012
COAL 9385,677.32976.13834.4324362662.85395815.7153.78132013
COAL 10386,1606353.8020.27166929263.25656016.9135.087132014
Cobalt 1127,104.564398.364608410,327.1159356.72665642046.793422042005
Cobalt 21,501,879.216307.999290.7614849.1765461748.6264042006
Cobalt 3584,388.8208.914,639.8301846.18964081723.8834042007
Cobalt 41,006,609.708253.991400934.03323461753.1664042008
Cobalt 5753,861.55239.00112,145.6116531.91864181874.3844142009
Cobalt 61,534,020.657.61911,888.5638841.227151122133.3914142010
Cobalt 72,423,752.548173.9769569.4736465.302755132516.4312142011
Cobalt 85,653,961.64194.51714.535.6171153.160152299.737142012
Cobalt 95,836,583.14318.92410,910.4391162.85395815.72327.204142013
Cobalt 105,843,888357.8081.51899E−0663.25656016.92044.3166142014
Copper 159,632.75812,949.277538226.78882556.7266564960.2797333052005
Copper 2704,626.1529619.83714006.95579949.176546820.3908052006
Copper 3274,173.610,137.48133893.48429646.1896408808.7823052007
Copper 4472,264.0261011.6532366.434.0332346822.5208052008
Copper 5353,683.9443571.844314.99759231.9186418879.3918152009
Copper 6719,705.74629.036159.43858641.227151121000.9083152010
Copper 71,137,135.0066554.6016079.8843665.302755131180.6164152011
Copper 82,652,629.586665.275880.22883453.160151078.9515152012
Copper 92,738,308.838284.5575344.36793462.85395815.71091.838152013
Copper 102,741,7368261.4199.67245E−0763.25656016.9959.1177152014
Gold 1563,683.544414,20335,480.7895556.72665649077.123076062005
Gold 26,660,536.59411,35418,934.3771949.1765467754.81144062006
Gold 32,591,648.48701872,636.9131346.18964087645.08114062007
Gold 44,464,114.50735799233.3987734.03323467774.94544062008
Gold 53,343,226.5394965.744931,652.965731.91864188312.52324162009
Gold 66,803,077.269619.837139,547.2310741.227151129461.16794162010
Gold 710,748,862.0712,949.2775326,051.5140565.3027551311,159.8735162011
Gold 825,074,199.04053,093.1312553.1601510,198.8777162012
Gold 925,884,089.1914,001.324144515.1029262.85395815.710,320.6884162013
Gold 1025,916,484.815,385.73892.03800558963.25656016.99066.13886162014
GRAPH 189,869.0863456275.874739756.72665641447.182133072005
GRAPH 21,061,901.3843567114.03235749.1765461236.3636072006
GRAPH 3413,191.23907162.805017846.18964081218.8691072007
GRAPH 4711,721.84219762149.4234.03323461239.5736072008
GRAPH 5533,016.6482463291.796699531.91864181325.2806172009
GRAPH 61,084,626.9741366.927125541.227151121508.4111172010
GRAPH 71,713,710.5027252494.981965.302755131779.2388172011
GRAPH 83,997,624.867022576.506978153.160151626.0255172012
GRAPH 94,126,747.116934537.550620162.85395815.71645.446172013
GRAPH 104,131,91268534.37076E−0863.25656016.91445.4309172014
IRID 134,435.818657.6787858.22449956.7266564554.5277333082005
IRID 2406,896.792687.9064229.46141749.176546473.7468082006
IRID 3158,325.6456.87054130.28155246.1896408467.0433082007
IRID 4272,715.846198.97824790.134.0332346474.9768082008
IRID 5204,240.024208.96999.14791831.9186418507.8178182009
IRID 6415,604.7253.9910,406.0632341.22715112577.9893182010
IRID 7656,655.426398.364608411,597.847465.30275513681.7644182011
IRID 81,531,800.18412.01421,823.5159353.16015623.0565182012
IRID 91,581,276.93519.60415,721.3688462.85395815.7630.498182013
IRID 101,583,256544.38821.88177E−0663.25656016.9553.8567182014
NICK 159,632.75820,789.80928,258.2346256.7266564960.2797333092005
NICK 2704,626.15221,677.09716,234.5864349.176546820.3908092006
NICK 3274,173.613,456.504719,105.2451946.1896408808.7823092007
NICK 4472,264.0263056.30815239.734.0332346822.5208092008
NICK 5353,683.9444857.52812,805.6873831.9186418879.3918192009
NICK 6719,705.76133.48318,132.6629641.227151121000.9083192010
NICK 71,137,135.0067992.18824,748.6424665.302755131180.6164192011
NICK 82,652,629.587898.71914,225.2191753.160151078.9515192012
NICK 92,738,308.8312,961.94711,243.4711362.85395815.71091.838192013
NICK 102,741,73616,632.7285.74789E−0663.25656016.9959.1177192014
PALA 195,468.40611,344.8915,030.3609756.72665641537.3492440102005
PALA 21,128,063.4649876.0786227.08812749.1765461313.39560102006
PALA 3438,935.27789.3055982.15292746.18964081294.81110102007
PALA 4756,065.8824678.70983585.434.03323461316.80560102008
PALA 5566,226.4085354.4497692.07886731.91864181407.85261102009
PALA 61,152,204.96916.10214,553.6530541.227151121602.39311102010
PALA 71,820,483.7428421.67465515,993.6502365.302755131890.09481102011
PALA 84,246,698.068136.21318,266.8714753.160151727.33551102012
PALA 94,383,865.319642.57520,269.3773662.85395815.71747.9661102013
PALA 104,389,35210,137.48133.0478E−0663.25656016.91535.48891102014
PHOS 1504,294.3568194.517102.759969656.726656494.675466670112005
PHOS 25,958,788.492173.97644.6337739249.17654680.88360112006
PHOS 32,318,594.74439.00147.9229009146.189640879.73910112007
PHOS 43,993,779.447048,679.682434.033234681.09360112008
PHOS 52,990,987.220031.918641886.70061112009
PHOS 66,086,311.20356.656104.751026541.2271511298.68111112010
PHOS 79,616,371.70146.0471015.79164665.30275513116.39881112011
PHOS 822,432,404.1516.79101.454979653.16015106.37551112012
PHOS 923,156,965.01061.3690163962.85395815.7107.6461112013
PHOS 1023,185,947.4400.09987774663.25656016.994.56091112014
PLATI 1504,294.3568599,079.3868,017.1546456.72665648120.7656120122005
PLATI 25,958,788.492408,475.8122,659.7037649.1765466937.7715320122006
PLATI 32,318,594.74435514223,390.3373846.18964086839.6023170122007
PLATI 43,993,779.447408,42219,27334.03323466955.784230122008
PLATI 52,990,987.226848.89934,909.1465731.91864187436.723321122009
PLATI 66,086,311.2038638.6547,352.9639541.227151128464.347861122010
PLATI 79,616,371.70110,826.5591439,646.182565.302755139984.079361122011
PLATI 822,432,404.1510,524.34544,137.9972153.160159124.333191122012
PLATI 923,156,965.0113,065.542401.7730562.85395815.79233.310021122013
PLATI 1023,185,947.4412,482.73411.1133E−0663.25656016.98110.938681122014
RHOD 178,026.52421319.9140,644.35656.72665641256.4786930132005
RHOD 2921,968.58481416.186,500.9839149.1765461073.440920132006
RHOD 3358,742.641256.8101,595.396246.18964081058.251770132007
RHOD 4617,934.1974509.728593634.03323461076.227920132008
RHOD 5462,778.0056568.13242,339.4123231.91864181150.640821132009
RHOD 6941,699.43726.90469,156.0302341.227151121309.639171132010
RHOD 71,487,885.099940.319985128,147.7710665.302755131544.778361132011
RHOD 83,470,835.042890.67635,053.1465553.160151411.754851132012
RHOD 93,582,942.1171146.08928,697.8775662.85395815.71428.61621132013
RHOD 103,587,426.41139.92351.24546E−0563.25656016.91254.958231132014
RHUT 127,324.68161345.0893839.07903756.7266564440.01550220142005
RHUT 2322,870.95041045.9832624.8671549.176546375.916160142006
RHUT 3125,630.72967.58045131.0656446.1896408370.596960142007
RHUT 4216,398.9152300.9871639.834.0332346376.892160142008
RHUT 5162,063.6288412.7771723.9041931.9186418402.951361142009
RHUT 6329,780.64555.0213637.41011641.22715112458.632161142010
RHUT 7521,053.4112823.03149143194.9055965.30275513540.977281142011
RHUT 81,215,477.216787.1692225.38954453.16015494.39281142012
RHUT 91,254,736.8161011.6531509.74459362.85395815.7500.29761142013
RHUT 101,256,307.2982.58046.96598E−0763.25656016.9439.483041142014
Appendix 2: Summary Statistics Summary statistics
VariableMeanMedianMinimumMaximum
fdi3.6338e+0061.2555e+0068399.02.5916e+007
out34637.3511.50.000005.9908e+005
price14964.4179.90.000002.8594e+005
vmi50.38051.15031.90065.300
cu50.90054.50034.00060.000
contr10.46010.0004.000016.900
labr2370.41302.580.00011160.
ps0.600001.00000.000001.0000
VariableStd. Dev.C.V.SkewnessEx. kurtosis
fdi5.9342e+0061.63312.61436.1548
out1.1316e+0053.26713.720712.776
price31417.2.09955.464739.860
vmi11.4720.22770−0.25720−1.2446
cu8.81160.17312−0.73808−0.93698
contr4.39730.420390.056813−1.4936
labr2803.91.18291.64811.4880
ps0.491660.81943−0.40825−1.8333
Variable5% Perc.95% Perc.IQ rangeMissing obs.
fdi59977.2.3121e+0073.5829e + 0060
out0.000004.0576e+0059149.70
price0.0000067271.16049.0
vmi31.90065.30021.7000
cu34.00060.00017.0000
contr4.000016.9009.00000
labr99.3509121.61766.00
ps0.000001.00001.00000
Appendix 3: Correlation Matrix Correlation coefficients, using the observations 1:01 - 14:10. 5% critical value (two-tailed) = 0.1660 for n = 140
outpricevmicucontrlabr
1.00000.00590.04370.0357−0.00270.3477
1.0000−0.1119−0.1358−0.15610.2088
1.00000.82240.55540.0728
1.00000.59130.0730
1.00000.0736
1.0000
Appendix 4 Model 1: Random-effects (GLS), using 140 observations; Included 14 cross-sectional units; Time-series length = 10; Dependent variable: l_fdi (b)
CoefficientStd. Errorzp-value
const8.250780.74680511.05<0.0001***
ps−1.036960.395350−2.6230.0087***
l_labr0.5104460.07179797.109<0.0001***
vmi−0.03932500.0166948−2.3560.0185**
cu0.02078840.02056901.0110.3122
contr0.3523660.04835617.287<0.0001***
Mean dependent var14.00240S.D. dependent var1.643830
Sum squared resid159.6645S.E. of regression1.087520
Log-likelihood−207.8516Akaike criterion427.7033
Schwarz criterion445.3531Hannan-Quinn434.8756

‘Between’ variance = 0

‘Within’ variance = 1.25397

theta used for quasi-demeaning = 0

Joint test on named regressors-

Asymptotic test statistic: Chi-square(5) = 181.228

with p-value = 2.92486e−037

Breusch-Pagan test-

ull hypothesis: Variance of the unit-specific error = 0

symptotic test statistic: Chi-square(1) = 3.70976

ith p-value = 0.0540952

Hausman test-

Null hypothesis: GLS estimates are consistent

Asymptotic test statistic: Chi-square(5) = 4.24038

with p-value = 0.515352

ReferencesKukaj, H. and Ahmeti, F.B. (2016) The Importance of Foreign Direct Investments in Economic Development in Transitional Countries: A Case Study of Kosovo. European Scientific Institute, ESI.Chimuka, L. (2007) Murowa Kimberlite Field: University of Zimbabwe Summer Symposium. University of Zimbabwe Publications, Harare.Dunning J.H. ,et al. (1988)The Eclectic Paradigm of International Production: A Restatement and Some Possible Extensions Journal of International Business Studies 19, 1-31.Demirhan, E. and Masca, M. (2008) Determinants of Foreign Direct Investment Flows to Developing Countries: A Cross-Sectional Analysis. Prague Economic Papers, 17, 356-369. https://doi.org/10.18267/j.pep.337Luiz, J. and Ruplal, M. (2010) Working Paper, 194. Wits Business School, University of Witwatersrand.Tsikita, G.K., Asante, Y. and Gyasi, E.M. (2000) Determinants of Foreign Direct Investment in Ghana. Overseas Development Institute.Anyanwu, J.C. (2011) Determinants of Foreign Direct Investment Inflows to Africa, 1980-2007. African Development Bank Group.Anyanwu J. ,et al. (2012)Why Does Foreign Direct Investment Go Where It Goes? New Evidence from African Countries Annals of Economic Finance 13, 425-462.Moyo, T. (2013) The Impact of Foreign Direct Investment on Economic Growth: The Case of Zimbabwe 2009-2012. International Journal of Economics, Finance and Management Sciences, 1, 323-329.Kariuki, C. (2015) The Determinants of Foreign Direct Investment in the African Union. Journal of Economics, Business and Management, 3, 346-351.Andreia Alexandra Faria Severiano (2011) The Determinants of FDI in Portugal—A Sectoral Approach. https://www.iseg.ulisboa.pt/aquila/getFile.do?method=getFile&fileId=503913Mody, A. and Srinivasan, K. (1998) U.S. and Japanese Investors: Do the March to the Same Tune? Canadian Journal of Economics, 31, 778-799. https://doi.org/10.2307/136491Bayraktar-Saglam, B. and Selin Sayek Boke, S.S. (2017) Labor Costs and Foreign Direct Investment: A Panel VAR Approach. MDPI, Open Access Journal, 5, 1-23.Lim, E. (2001) Determinants of, and the Relation between, Foreign Direct Investment and Growth: A Summary of the Recent Literature. International Monetary Fund Working Paper, Middle Eastern Department.Varamini, H. and Vu, A. (2007) Foreign Direct Investment in Vietnam and Its Impact on Economic Growth. International Journal of Business Research, 7, 132-139. http://www.etsg.org/ETSG2009/papers/forte.pdfDjokoto, J.G. and Dzeha, G.C. (2012) Determinants and Effects of Foreign Direct Investment in Ghana—Review of Literature. The International Institute for Science, Technology and Education, 2, 25-35.Aswal, N. (2014) China’s Integration into Global Trading System and Its Implications for India. Doctoral Thesis.Khurtish-Kastrati S. ,et al. (2013)The Effects of Foreign Direct Investments for HostCountry’s Economy European Journal of Interdisciplinary Studies 5, 26-38.Hansen, H. and Rand, J. (2006) On the Causal Links between FDI and Growth in Developing Countries. World Economy, 29, 21-41. https://doi.org/10.1111/j.1467-9701.2006.00756.xAdewuni, S. (2006) The Impact of FDI on Growth in Developing Countries: An African Experience 1970-2003. Jonkoping International Business School, Sweden.Oluwatocin, A., Olusegun, O., Festus, O.E. and Abibola, O. (2012) Foreign Direct Investment, Economic Growth and Financial Sector Development in Small Open Developing Economies. Economic Analysis and Policy, 42, 105-127. https://doi.org/10.1016/S0313-5926(12)50008-1Javorvick, B. (2004) Does Foreign Direct Investment Increase the Productivity of Domestic Firms? In Search of Spillovers through Backwards Linkages. American Economic Review, 94, 605-627. https://doi.org/10.1257/0002828041464605Hsiao, C. (1986) Analysis of Panel Data, Vol. 11 of Econometric Society Monographs. Cambridge University Press, New York.Henry, J.F. (2000) Adam Smith and the Theory of Value. Department of Economics, California State University, Sacramento.Dunning, J.H. (1993) Multinational Enterprise in the Global Economy. Addison Wesley Publishing Co., Wokingham.Ramirez, M. (2010) Economic and Institutional Determinants of FDI Flows to Latin America: A Panel Study. Trinity College, Department of Economics.Fasanya, I. (2012) Capital Flows-Growth Nexus in Nigeria: Has Foreign Direct Investment Played a Role in Accelerating Economic Growth? Journal of Sustainable Development in Africa, 14, 34-52.Hermes, N. and Lensink, R. (2003) FDI, Financial Development and Economic Growth. Journal of Development Studies, 40, 142-163. https://doi.org/10.1080/00220380412331293707Buckley, P.J., Clegg, J. and Zheng, P. (2007) The Impact of Foreign Direct Investment on the Productivity of China’s Automotive Industry. Management International Review, 47, 707. https://doi.org/10.1007/s11575-007-0041-1Ikechi, K.S., Emmanuel, N.C., Onyechere, P.O. and Sussan, O.A. (2017) Foreign Direct Investments and Industrial Productivity in Nigeria. Journal of Economics and Sustainable Development, 8, 162-180.Orji, A., Antoner-Orji, O.I., Nchege, J.E. and Okarfor, J. (2015) Manufacturing Output and Foreign Direct Investment in Nigeria: A New Evidence. International Journal of Academic Research in Economics and Management Sciences, 4, 16-28.Oladipo O.S. ,et al. (2010)Foreign Direct Investment (FDI): Determinants and Growth Effects in a Small Open Economy The International Journal of Business and Finance Research 4, 75-88.Mulenga, M.C. (2012) An Investigation of the Determinants of Intra-Industry Trade between Zambia and Its Trading Partners in the Southern African Development Community (SADC). TIPS Working Paper Series.Krugman, P. and Obstfeld, M. (2009) International Economics: Theory and Policy. 8th Edition, Pearson Addison Wesley, Boston.Hussain, F. and Kimuli, C.K. (2012) Determinants of Foreign Direct Investment Flows to Developing Countries. SBP Research Bulletin, 8, 13-31.Chamber of Mines Annual Reports (2012, 2013) Harare.