DETERMINANTS OF GHANA’S ECONOMIC GROWTH BASED ON FACTOR ANALYSIS.
Isaac Okyere Paintsil
Email: Tel: 008615962044635
School of Finance and Economics, Jiangsu University. Zhenjiang, China PRC.
Abstract
Ghana’s economy achieved commendable growth in 2014 when it reached a millennium development goal (MDG) of halving poverty. The challenge now is how the economy can maintain sustainable growth in the coming years. We use exploratory factor analysis procedures to determine two important factors that drive the economy namely; Income generation, and Production efficiency. We realized that Income generation contributes a greater part of the variance in growth and concluded that policy makers in should deepen their commitment to implement policies that would spark income growth and productivity.
Keywords: Economic growth, Factor analysis, Parallel analysis.
1. Introduction
The economy of Ghana like most other economies grows at varying rates each year and faces various challenges such as currency depreciation, energy crisis, macroeconomic imbalance, inflation, trade imbalance etc. However, the country has over the past few years made gains in accelerated economic growth which saw the country achieving her millennium economic goal (MDG) of halving poverty even though growth disparities still exist among the regions of the country.
Research concerning economic growth is very popular among economists and various theories and models have been popularized to ensure policy makers take correct measures to promote growth. However, developed economies have enjoyed a greater share of these studies using various econometric methods. Consequently, when both developed and developing economies are studied together, findings mostly show disparities in results.
This paper focus on determining the underlying factors that drives economic growth in Ghana using exploratory factor analysis method. Factor analysis is basically about investigating whether a number of variables of interest are linearly related to a smaller number of unobservable factors.
The rest of the paper is organized as follows: the section 1 is introduction of the study. Section 2 deals with literature review on related research theories concerning the topic of this paper. Section 3 is about identifying the various indexes to be used in the analysis. Section 4 is about performing the step by step factor analysis. Section 5 shows results from our analysis. Section 6 is about conclusion and suggestions based on the outcome of the study.
2. Literature review
Economic growth is central to the development planning of every region and country. Policy makers and development planners aim to identify the various sources that instigate economic growth in their economies in order to effectively direct policies to achieve growth and development goals. Consequently, the explanatory variables for economic growth have been popularly studied and many models have been used to study them. One of such models is the neoclassical model.
Solow (1956) pioneered the neoclassical growth frame work by allowing for substitutability in order to achieve growth stability. The neoclassical framework set in motion many subsequent extensions and applications over the last decades. Particularly; Sidrauski (1967) included money and inflation; Brock and Mirman (1972) studied the neoclassical model with uncertainty; Blanchard (1985) analyzed the impact of government spending, debt, and deficits in the model; Barro (1990) studied the implications of government spending; Mankiw, Romer, and Weil (1992), introduced human capital as a third factor of production in an attempt to reconcile the neoclassical model with existing evidence on convergence rates. Similarly, Caselli and Ventura (2000) included various forms of household heterogeneity within the Cass-Koopmans-Ramsey model, taking cue from Laibson’s (1997) insights on hyperbolic time discounting, and Barro’s (1999a) neoclassical model with non-constant time-preference rates.
One major drawback in Solow’s (1956) model was that growth per capita output eventually converged to zero in the stable state. He suggested that in order for a steady state to occur, exogenous technology must be introduced. There were other empirical findings that suggested problems with Solow’s model. One suggestion of his model is that economies with similar technologies and preferences will converge to the same steady state output levels, however, findings by De Long (1988), Quah (1989), and Romer (1989c) showed that here is very little evidence of convergence for a broader sample of economies.
Given the difficulties in Solow’s model with the Solow model, many growth theorists have proposed new models that attempts to endogenize the processes of economic growth. Rebelo (1987) presented a simple model in which capital is linearly related to output which expressed the production function in the form: where is a composite of physical and human capital. Here it is easy to show that sustained per capita output growth is realistic without dependence on exogenous technical change. This came to be known as the model which assumes that as people accumulate capital, and learning through doing, technological progress would be stimulated that would eventually raise the marginal product of capital, thus eliminating the tendency for the marginal product to diminish when technology is unchanged. Subsequent extensions and applications of this model based on this model have also been undertaken in recent past years.
Barro (2003) studied the determinants of economic growth in a panel of countries using data from 1965 to 1995. He included explanatory variables such as per capita income, educational attainment, life expectancy, fertility rate, Government consumption ratio, rule of law, democracy, international openness, the terms of trade, investment ratio, and inflation rate. His findings were consistent with the neoclassical model. In his conclusions he noted that high initial human capital predicts per capita growth. Also a given per capita and human capital depends positively on rule of law and international openness, and negatively on inflation rate and ratio of government consumption to GDP. Similarly growth increases with favorable terms of trade and declines with soaring fertility rate. Additionally, growth and investment ratio relates positively but weakens when all other variables are held constant.
Aryeetey et al. (2001) in an effort to study the sources of development in Ghana adopted Cobb-Douglas production model using data on labour growth, capital growth, GDP growth, and trade liberalization from 1961 to 1996. Their findings indicated that the only significant variable is trade liberalization which had a positive coefficient and concluded that other factors not included in their research could account for economic growth.
Edwin et al. (2012) as used the Augmented Cobb-Douglas production function to study the main sources of economic growth in Ghana using data from 1991 to 2011. They concluded that capital, labour, and total factor productivity were the main drivers of Ghana’s economic growth.
3. Construction of Economic Growth Evaluation Index
Due to the broad nature of the indicators of economic growth, availability of data, measure of comprehensiveness, and based on related researches on the subject matter such as those captured in the literature review section, we select the following indicators to evaluate their contribution to economic development in Ghana. The data was collected from the World Bank time series data on Ghana from 1991 to 2011.
3.1. Gross Domestic Savings. ()
Savings accumulation can be considered to be an important source of capital stock which can be drawn to inject into investments. Consequently, Gross Domestic Savings is one of the important factors contributing to economic growth (Mankiw, 2008).
3.2. Tax rate. ()
The relationship between tax rates and economic development appear, according to some researchers, perineal and sometimes controversial. It is believed that the effects of revenues from the corporate income tax and personal income tax separately are sensitive to specification, but when they are significant, they are often positive, indicating that higher taxes and greater dependence on these taxes are associated with faster growth (Alm and Rogers, 2011). Similarly, Tomljanovich (2004), using data from 1972 to 1998, realized that higher taxes reduce short-term growth rates. However, the short-term reduction in growth rates does permanently affect the size of the economy. After he decomposed total tax burden into components, he noted that income, property, and sales taxes have no significant effects, but corporate taxes have positive effects on growth.
3.3. Stock Markets ()
Stock markets enhance economic activities since in they help in generating liquidity in a short term. According to common theory the price of a share is equivalent to the sum of all future dividend payments discounted to its present value. Higher stock prices indicate a rise in the discounted expected earnings, which provides useful indication about future economic growth. Higher stock prices also offer an extra economic impetus to individuals and companies that own. Similarly, the stock market is usually used as an indicator of the state of the economy through which stock prices affect the real economy through a confidence channel. An increase in stock prices would increase confidence of individuals, households and firms and also minimize their uncertainties concerning their future economic situation.
3.4. Foreign Direct Investment () and Income from FDI ()
As earlier discussed in the literature review section it is often the case that FDI holds a huge impact on economic growth especially in developing economies. On the other hand the impact of FDI is sometimes heterogeneous depending on the type of FDI, the absorption capacity of the hosting economy and the economic sector. Spillovers from FDI contribute to local industrial growth, however, in terms of completion, FDI impact often tends to be negative.
3.5. Energy Consumption ()
Energy consumption and economic growth are often believed to be related, but the direction of causality is still not clear. There is no consensus as to whether energy consumption causes growth or growth increases energy consumption. However, we include energy consumption in our study due to the crucial role it plays in industrial growth.
3.6. Cash crop production ()
Ghana is one of the world’s leading producers of cocoa which has a great value on the international markets. Over 60% of Ghana’s population are involved in agricultural activities especially cash crop production. It is expected that export of cash crops would contribute to the economic growth.
3.7. Infrastructural Development () and Urbanization ().
The advantages of infrastructural development cannot be overemphasized, apart from improving the general well-being of people, it also helps to attract investment and promote development. McCosky and Kao (1999) studied the impact of urbanization on economic growth in industrialized and developing countries using time series data. They discovered the two variables are cointegrated but the nature of relationship varied across countries.
3.8. Monetary Policies (Primary Income (), Quasi Money () , and Total Reserve (), Government Policy rate())
Monetary policies are very crucial to economic growth in every economy. Exogenous growth literature pioneered by Romer (1986) has helped to understand the impact of policies to economic growth. Chari, Jones, and Manuelli (1995), and Van der Ploeg and Alogoskoufis (1994) studied the impact of monetary policies on long run real activity. Their findings on how for instance affect growth helped in advancing knowledge that monetary policies have effect on growth.
3.9. Labour Force Participation ().
Labour force growth presents opportunities for economic expansion. Labor force growth is determined by increase in the births, net immigration, and the labor force participation rate. Increased labour force participation provides the requisite human resources needed for production of goods and services.
3.10. Innovation ().
The crucial role innovation plays in economic growth has attracted many researches in the past decades. For example Solow (1957), studying the US productivity between 1909 and 1949, concluded that the main factor responsible for the majority of the economic growth (87.5%), during that period was the technical change. Similarly, studies by Matthews et al, 1982; Denison, 1985; Jorgenson, 1990 suggest that technical progress is responsible for an important part of economy.
3.11. Trade ().
Trade allows an economy to make optimum use of its resources and ensure increase in output. Trade liberalization stimulates economic growth and efficiency by allowing producers to exploit areas in which they have a comparative advantage over foreign producers and by reducing their real costs.
4. Factor Analysis
We conduct a Kaiser-Meyer-Olkin (KMO) and Bartlett test to examine the variance among variables to determine if they are interdependent and suitable for factor analysis. The value range of KMO test given by Kaiser (1974) should be (0, 1). A KMO test results > 0.7 is acceptable and indicate that the variables are measuring a common factor. The value of our KMO test is 0.79, meaning conducting factor analysis is feasible. Our Bartlett’s test p-value is approximately which is less than 0.05 indicating that our sample variables are suitable for analysis.
Table 1. KMO and Bartlett TestKaiser-Meyer-Olkin (KMO) measure of sampling adequacy / 0.79
Approx. chi-square / 679.0538
Bartlett’s test of sphericity / Df / 120
P-value / 6.419166e-79 ()
4.1 Factor Extraction
By the maximum likelihood approach and a scree plot from parallel analysis, we proceed to determine the number of underlying factors to be used in our analysis. Based on the Kaiser criterion which states that factors with eigenvalues greater than 1 are significant and stable we extract two factors for the analysis. Table 2 shows two factors with eigenvalues greater than 1, implying that two factors could explain the original 16 variables to explain economic growth in Ghana.
Table 2. Characteristics variance contributionFactor 1 / Factor 2
Proportion of variance / 11.7756 / 1.6188
Similarly, by adopting the maximum likelihood method we observe the relationship between each factor and the variables and we use oblique rotation for the factor loadings. In the Oblique solution we consider eigenvalues greater than 0.3 (ie greater than 10% of variance between factors) as a criteria for loadings to obtain a simple structure. Each factor that shows 0.3 or greater loading has a greater reliability. We eliminate variables that double load or does not load on any of the factors. Consequently we eliminate innovation from our sample variables because it loads on factor 1 and factor 2.