Embedded Files
Sokroeun Rith
August 06, 2019
Submitted in Satisfaction of the Requirements of
Economics 145 – Economic Research Methods
California State University
Department of Economics
Professor Dowell
Summer 2019
What are the Determinants of Economic Growth?
Abstract
Empirical findings for a panel of 13 developed countries from 1998 to 2016 strongly support the general notion of conditional convergence. For a given starting level of real GDP per capita growth, the growth rate is enhanced by the decrease in labor force participation and patent applications of residents, as well as lower inflation. Additionally, growth is stimulated by the increase in human capital, physical capital, population, and improvements in terms of trade openness. However, research and development expenditure (R&D) is found to have a negative coefficient and is statistically insignificant.
JEL Codes: I20, J00, J24, O11, O30, O40
I. Introduction
Economic growth is crucial because it is a vital component of a complex chain of reactions that positively affects a country's economy (Samuelson and Nordhaus, 2011). The most prominent motivator for these events is an improved standard of living for the citizens of the country. When an economy expands, the Gross Domestic Product (GDP) also increases in proportion. In simple terms, GDP is the monetary sum of all the finished goods and services that the particular economy produced in that corresponding year.
When GDP increases, GDP per capita also increases, meaning that each individual person experiences an increase in net income per year. An increased GDP per capita allows people to have the funds necessary to reinvest in the market and continue the circulation of currency so that they can spend more, save more, and afford not only necessities but also luxury items. Therefore, an important question arises regarding the factors that enhance economic growth. What are the determinants of economic growth?
Keeping this in mind, I will focus in this paper on analyzing the relationship between some of the economic variables on output growth, using the general growth theory. The endogenous growth model developed by Romer (1990) asserts that labor, ideas, and human capital contribute to output. In this research paper, I have chosen 13 developed countries to examine their economic conditions using panel data from 1998 to 2016. The reason for selecting only developed countries for this paper is that they will have enough and accurate data compared to developing countries.
Keeping this in mind, I will focus in this paper on analyzing the relationship between some of the economic variables on output growth, using the general growth theory. The endogenous growth model developed by Romer (1990) asserts that labor, ideas, and human capital contribute to output. In this research paper, I have chosen 13 developed countries to examine their economic conditions using panel data from 1998 to 2016. The reason for selecting only developed countries for this paper is that they will have enough and accurate data compared to developing countries.
The dependent variable in my analysis is real GDP per capita growth (RGDPg), while my independent variables are labor, patent, research and development expenditure (R&D), gross fixed capital formation (GFCF), trade openness, inflation, population, and tertiary education. The data used in this analysis were retrieved from three different sources, including the World Bank, FRED, and OECD.org.
II. Literature Review
Silaghi, Pop, and Medesfalean's (2014) insights about the determinants of economic growth are closely related to my paper. However, their paper introduces the Solow (1956) growth model, while I use the endogenous growth theory, Romer (1990). The empirical results in their paper indicate that capital and labor are positive coefficients, patents are found to be negative and significant, while human capital is insignificant. The gap that I found in their paper is that they only use one country to examine the determinants of growth, while my paper collects panel data from 13 developed countries.
Barro's (1996) paper on the determinants of economic growth finds that the growth rate level of real GDP per capita is enhanced by human capital. He also believes that there is a link between human capital and physical capital. Physical capital and human capital together contribute to the economic growth of countries because the effective use of physical capital is possible only with a skilled and educated workforce.
However, Pritchett (1996) finds that human capital has a negative and significant coefficient for economic growth. This indicates that in some cases, the education system may have failed to adequately prepare individuals, or human energy may be directed towards unnecessary areas. Other studies, such as those by Benhabib and Spiegel (1994) and Engelbrecht (2001), also find negative and insignificant coefficients for human capital.
Empirical results from Barro (1996) also indicate that lower inflation enhances real GDP per capita growth. This makes sense because when inflation increases, the purchasing power of individuals decreases since the prices of goods and services increase. Therefore, the lower the inflation rate, the higher the growth rate of real GDP per capita will be.
The relationship between population and economic growth is discussed in Oaka's study (2008), which finds that there is a strong correlation between population and real GDP per capita, supporting the hypothesis that population drives economic growth. However, Tsen and Furuoka's study (2005) finds no such link between population and economic growth.
Does trade openness matter for economic growth in the CEE Countries from Iyke (2017). He says that the increase in trade openness (import and export) will lead to an increase in GDP per capita. The empirical results find that trade openness contributes to economic growth in 17 central and Eastern Europe countries, and it is one of the best predictors for economic growth. This paper is similar to my paper since he collects panel data from many countries in Europe to examine the key macroeconomic variables to find the relationship between these variables.
Patent rights and economic growth are from Hu and Png (2012). This paper talks about patents that impact economic growth. To promote these new ideas, governments offer patents to protect the individual’s ideas and inventions for a specific amount of time. These patents eliminate potential direct competition until it expires, thus giving the patent holder monopolistic control over their idea or product for time being. However, it can be argued that patents also limit economic growth because they place restrictions on the dispersion and alteration of the patented idea.
Research done by Wennekers and Thurik (1999) have demonstrated that there is a positive link between patents and economic growth in the long-run even though there is an initial negative correlation in the short-run. Other studies have shown that there is a negative correlation between GDP trends and patent application trends in economic growth in poorer economies (Saini and Jain 2011).
Despite this, research done by Grossman and Helpman (1990) has shown that innovation, and patents help consumer demand and competition. It has been shown that there
is a positive correlation between research and development (R & D) expenses and patents because ideas are usually not patented unless they have been thoroughly researched,
tested, and proven to be effective and positively accepted. Also, the empirical study shows that R&D expenditure is statistically significant and positive coefficient with
economic growth. This result makes sense though since R&D is one of the main factors for economic growth in both developing and developed countries (Sahin 2015).
III. Economic Theory and Model
Romer (1990) the new economic growth model. This theory is different from the Solow model. The fundamental distinction of this theory is to divide the world of economic goods into objects and ideas which lead to a sustained growth, whereas the Solow model fails to do so because it focuses only on the capital accumulation as a possible engine that leads towards economic growth. Objects are something like goods. For example, cellphones, jet planes, computers, and papers, as well as labor and capital from the Solow model. Ideas are instructions or recipes that make objects. Ideas, for example, are the set of instructions for changing trees into paper, turning petroleum into plastic, and converting sand into computer chips. Following are the assumptions of the theory:
People and the existing stock of ideas contribute to output
New ideas are created by the stock of existing ideas and the number of workers.
Workers produce ideas because they engage in research.
According to the Macroeconomic textbook, Jones (2014), the production function is:
Yt=AtLyt (1).
∆At+1=zAtLat (2).
The first equation (1) is the product function of output Yt. Output is produced by labor Lyt and the pervious ideas At.
The second equation is (2) the production function for new ideas. At is the stock of ideas at time t. is the “change over time.” At+1 is the number of new ideas produced during period t. The second equation says that new ideas are produced by using existing stock of ideas and workers. The only difference between two production functions is that the second one includes a productivity parameter z. z is how efficient researchers are at producing new ideas.
Lat+Lyt=L (3).
Equation (3) is what we call resource constraint. The fraction of workers produce ideas is
Lat and Lyt is the number workers produce output add up to the total population L.
Lat=lL , Lyt=1-lL (4).
From the equation (3) above, we know only Lat is the fraction of workers that produces new ideas by engaging in research, while Lyt is the number workers produce output, but we do not know how labor is allocated, or how much labor produces output. Equation (4) is the allocation of labor. Through a rule of thumb, we assume that l is the constant fraction of population, leaving that 1-l workers produce output. We can set l = 0.05. Thus, 5 percent of the workers produce new ideas while 95 percent of workers produce good consumption.
IV. Method Analysis
Given that this paper will be focused on investigating the relationship between independent variables and growth. Real GDP per capita growth is my dependent variable, and independent variables will be used for this analysis are illustrated in the following table:
All of the independent variables listed above are dictated by the theory and significant in the existing literature. Labor is considered to be positive. Solow (1956) and Romer (1990) the new economic growth model, labor is one of the production functions to produce output. Labor contributes to economic growth because it represents the human factor that produces goods and services in the economy. China, for example, because it has a huge amount of labor force can stimulate its economy to grow rapidly.
Patents and research and development (R&D) also believe to be positive. To promote the new ideas, governments offer patents to protect the individual’s ideas and inventions for a specific amount of time. Romer’s theory says that the ideas contribute to sustained growth. R&D is often an indication of a country’s dedication to science and technology. Many studies have shown that R&D has positively effect's factor productivity by increasing output per worker and on GDP per capita.
Gross fixed capital defined as physical capital believes to be positive because physical capital is a part of the production function in the Romer model. Economists call it a factor of production. It includes things like machinery, computers, equipment, and buildings. These things help turn the raw materials into the finished services and products. Moreover, equipment and other physical things a business owner invests money into when they want to produce something.
The more free trade, the more competition increases. This will allow the world prices to decrease, which provides benefits to consumers by raising the power of their income, and leads to a rise in consumer surplus. Freeing trade also allows for access to new ideas and capital that not only boosts growth via input A in Romer’s model but also serves to lessen the trade gap distinguished by Romer (1990, 1986) and Barro (1996).
Barro (1996) shows that lower inflation will enhance the growth rate of real GDP per capita. As we know that when prices increase, it will impact living expenses, the cost of doing business, mortgage, borrowing money, and every other face of the economy. In contrast, if inflation decreases, this will allow people to borrow more money and expand their businesses so that this will cause the economy to grow.
Population considers to be positive. As we know in the Romer’s model labor produces output. Thus, the more population the more labor force we have and the more output will be produced. For example, China and India, as both have more populations, their economy grows rapidly since they have more workers to produce output.
Tertiary school enrollment defined as human capital will have a positive sign. When more people enroll in tertiary school, this means they are going to become what we call human capital. Thus, the more human capital we have, the more productive they will be. This allows for technologies to be readily created, implemented, and adopted in the economy, which in turn stimulates economic growth.
Following presents a simple growth model that attempts to capture the effect of some of the economic policy variables on output growth.
GDPg= 0+1LABORit+2PATENTit+3R&Dit+4GFCFit+5TRADEit+6INFLATIONit+7POPit+8EDUit+uit (5).
The explanatory variables chosen in the above equation (5) are in the growth regression from Barro (1991) and Easterly (1993), as well as others that are common in the literature.
V. Empirical Methodology
A. Data
To conduct my analysis, I used panel data that is retrieved from multiple sources including World Bank, FREED, and OEOD.org. The world bank was the data source for the dependent variable which is real GDP per capita growth, and independent variables are labor, patent, R&D, trade openness, and gross fixed capital formation. Inflation and total population were from the FREED. The last one was tertiary education which was from OECD.org.
B. Descriptive Statistics
Table #3: Summary statistics from 1998-2016
Table 3 above gives the summary statistics for the pooled sample from 1998 to 2016 in 13 developed countries. The mean value compared to the
median for every variable is reasonable with the exception that the three variables are patent, trade, and population. The mean is way higher than
its value. The value of standard deviation of all variables seems to be good as they are small compared to the mean except the patent and
population, but this does not surprise since a few of those 13 developed countries contain a small amount of population. The mean value for real
GDP per capita growth compared to the min and max. Seeing that the min value is -14.560, while the max value is 10.924. Thus, we can interpret
that the value of decreasing in real GDP per capita is greater than the value of increasing.
As we can see from table 4 above, the correlation matrix, the linear correlation between the dependent variable (RGDPg) and labor, is relatively small but positive with a value of 0.228788. This weak correlation does not reflect the Romer model that labor produces output. The other correlations that surprise are those between patents, R&D, and tertiary education are relatively small and negative with respect to RGDPg. These independent variables show that they do not have any relationship with RGDPg, which does not reflect the theory. GFCF, trade openness, and inflation are positively correlated to RGDPg. However, these independent variables seem to have a weak relationship because the number as we can see from the above is not close to 1. Population is also relatively small and negative to RGDPg with a value of -0.10549. This does not reflect the theory as well.
The graph above shows the average of real GDP per capita growth over a 19-year period from 1998 to 2016 of 13 developed countries. If we look at the trend, there was a peak in 2011, and then went down during the period of financial crisis which was in 2008 to 2009. It then went back up again as the economy improved.
The graph above shows the labor force participation over a 19-year period from 1998 to 2016 in 13 developed countries. Over time the trend increased in the period. However, there was a little decrease in 2001.
The graph above shows the patent application of residents over a 19-year period from 1998 to 2016 of 13 developed countries. This graph seems to have a relationship to the RGDPg because the patent went down at the same period as the RGDPg was. However, in the correlation matrix, the patent shows that it is relatively small and negative to RGDPg.
In figure 4 above, the regression line looks like a downward slope, and it does not fit the observation points between RGDPg and labor. The observation points are all over the place. This means that the RGDPg and labor have a negative correlation. However, in the correlation matrix in table 4, labor has a positive sign to RGDPg which is 0.22879.
Patent and RGDPg are not related to each other as we can see in the figure 5 above. The regression line does not fit the observation points. The observation points are all over the place while some are missing in the middle of the line. This means that there is a strong negative relationship between RGDPg and patent.
V. Statistical Analysis
The following regression results are based on a sample of 13 developed countries from 1998 to 2016 with 247 observations, but after inflation and population are lagged once the observations become 234. This paper is also organized as panel data so that the following regressions are reported with robust standard errors, Ordinary Least Squares (OLD), fixed effects, and stationarity.
We think it is important to run a stationarity (unit root) test for our paper because we can expect the results. For example, after running the unit root test, if we see that the p-values are higer than 0.05, then that means the values are non-stationarity and that we cannot expect good results. After running the unit root test, we found that real GDP per capita growth (RGDPg), LABOR, PATENT, EDU, GFCF, TRADE, INFLATION, and POP are all stationarity except R&D is non-stationarity, which the p-value is 0.4849.
We also did Heteroscedasticity and Fixed Effects (redundancy) tests. After we had done the Heteroscedasticity test, we found both cross-section and period with a low p-value, which means we reject the null, meaning there is heteroscedasticity.The test for redundancy, both the cross-section and period fixed effects, cross-section fixed effects, and period fixed effects, indicates that there is no redundancy as the p-value is less than 0.05.
The following table shows the summarized Ordinary Least Squares (OLS) regression results for the balanced panel.
Table 4 above shows the results that the independent variable of LABOR is significant with respect to the dependent variable RGDPg as suggested by the lower p-values across all four regressions are lower than 0.05. Thus, these results lead to the conclusion that the null hypothesis should be rejected. However, all four regressions have a negative estimated coefficient sign, which means an increase in labor force participation tends to reduce the growth rate of real GDP per capita. This result does not reflect the economic theory and the empirical results found in the literature.
Patent is also shown to be significant in four regressions, but the estimated coefficient is negative in all four regression. This concludes that an increase in patent application of residents tends to decrease the growth rate of GDP per capita. R&D displays all four regressions as statistically insignificant, which means there is insufficient evidence to conclude that a non-zero correlation exists. These results contrast with the empirical results found in the literature. Yet, it was probably because of the data errors. EDU is defined as human capital. EDU is shown to be significant in both regression 2 and 3, but there is a negative sign in regression 2 and 4. This is different from the economic theory and what we expected that human capital should have a positive coefficient sign. The sign and significance of the independent variable GFCF is consistent across all four regressions as predicted in the Romer model, as well as found in the literature. The growth rate of GFCF will cause the real GDP per capita to grow. Trade is shown to be significant in three regressions 2, 3, and 4 with a positive coefficient sign. This means an improvement in the terms of trade does stimulate an expansion of domestic products. The one-period lag of INFLATION is shown to be significant in regression 1, 2, and 4 with a negative sign across four regressions. This interprets that the decrease in inflation will increase the growth rate of real GDP per capita. These two variables have a reverse relationship. The last one independent variable to be discussing POP. The one-period lag of POP is shown to be statistically significant and positive coefficient in all four regressions. Thus, an increase in population will improve the economy and increase the growth rate of real GDP per capita.
In the last section of the table R2, R2, and P-value are reported. The higher the R2, the better the model fits the data. R2 is significant in all four regressions as we can see. The value of R2 In regression 1 is low if compared to the other three regressions. R2 value is consistent with the R2 value with the similar pattern in each regression result. P-values of all regressions are low. Thus, this tells us that it is significant.
VI. Conclusion
The question addressed in this paper is what are the determinants of economic growth in recipient countries. Upon analyzing a panel of 13 developed countries from 1998 to 2016 with the OLS econometric technique, in all the specifications and combinations of fixed effects, the evidence indicates that the growth rate of real GDP per capita is enhanced by the decrease of labor force participation (LABOR), patent application of residents(PATENT), and lower inflation.
Growth is also stimulated by the increase in human capital (EDU), physical capital (GFCF), Population (POP), and by improvements in terms of trade. Research and Development expenditure (R&D) indicates to be insignificant and negative coefficient across all four regressions. These results confirm findings in the literature, as well as the economic growth theory except R&D is found to be different from the literature.
References
Samuelson, P and Nordhaus, W. 2001. Economics, Bucuresti, Editura Teora
Wennerkers, S and Thurik, R. 1999. “Linking Entrepreneurship and Economic Growth” Small
Business Economics 13, pp.27-55
Hu, A. and Png, I. (2012). “Patent Rights and Economic Growth: Evidence from Cross- Country. Panels of Manufacturing Industries," CELS 2009 4th Annual Conference on Empirical Legal Studies Pape
Silaghi, Monica Pop, and Romona Medesfalean. 2014. “Some Insights about Determinants of Economic Growth in Romania: An Empirical Exercise.” Theoretical and Applied Economic 21 (6): 23-36.
Wennekers, S. and Thurik, R. (1999). “Linking Entrepreneurship and Economic Growth.”
Small Business Economics 13, pp. 27-55
Grossman, G. and Helpman, E. (1990). “Trade, Knowledge Spillovers and Growth," NBER
Working Paper No.3285
Barro, Robert. "Determinants of Economic Growth: A Cross-Country Empirical Study." 1996. doi:10.3386/w5698.
Benhabib, J and Spiegel, M. 1994. “The Role of Human Capital in Economic Development Evidence from Aggregate Cross-Country Data”, Journal of Monetary Economics, 34
Engelbrecht, H. 2001. “The Role of Human Capital in Economic Growth: Some Empirical Evidence on the “Lucas vs. Nelson-Phelps” Controversy”, Massey University, Discussion Paper No. 01.02
Pritchett, L. 1996. “Where has all the Education Gone?”, Policy Research Working Paper 1581, The World Bank
Easterly, W and Rebelo, S. 1993. “Fiscal Policy and Economic Growth: An Empirical Investigation.” Journal of Monetary Economics 32: 417-458
Barro, R.J. 1991. “ Economic Growth in a Cross-section of Countries.” Quarterly Journal of Economics 106: 2 407-443.
Romer, P.M. 1990. “Human Capital and Growth.” Paper Presented at the Carnegie Rochester Conference on Economic Policy, Rochester, New York.
Romer, P.M. 1986. “Increasing Returns and Long-Run Growth.” Journal of Political Economy 94:5 1002-1037
Lucas, R. 1988. “On the Mechanics of Economic Development: W.A. Mackintosh Lecture.” Journal of Monetary Economics 22: 1 3-42
Barro, R.J. and Lee, J.W. 1994. “Losers and Winners in Economic Growth.” Proceedings of the World Bank Annual Conference on Development Economics, 267-297
Psacharopoulos, G. 1993. Returns to Education. San Francisco: Jossey-Bass
Page updated
Google Sites
Report abuse