Panel Data Analysis of Human Development, Income Inequality and Economic Growth in the SAARC Countries During 1990-2020

The present study is undertaken to analyze human development, income inequality and economic growth in the SAARC countries, along with china, during 1990 to 2020 using dynamic panel data models. The Hausman test was used to choose between Fixed Effect (FE) or Random Effect (RE) models. The major findings include that income inequality (G) negatively affect human development (HD), whereas, economic growth (EG), Gender Development Index (GDI) and mean year of schooling (MYS) positively affect human development (HD). On the other hand, control of corruption (CC), G and HD accelerate EG, and population growth per annum (P) slow down the rate of EG in the panel of countries. Moreover, EG significantly increases the G, and agriculture area as percentage of total land area (AGL) and HD negatively and significantly affect income inequality (G) in the sample of countries. Summing up, human development positively and significantly accelerates economic growth, and economic growth enhance the level of human development ranking of the sample countries. Moreover, human development measure also significantly reduces income inequality.


Introduction
People are both means and end of economic development, (Haq, 1995).In the words of Ranis (2004), human development is the ultimate goal of development processes or putting it differently, Ravallion (1997) stated that human development is the end and economic growth is the means.However, Investment capital or physical capital is given the due importance while discussing means of development, but human capital is ignored or given less attention than it deserves.Human capital which includes human institutions and skills are vital to consider in making real development, (Haq, 1995).In the words of Alkire (2002), human development is the flourishing of human beings in its fullest sense.It includes flourishing in matters related to public and private, economic, social, political and spiritual matters.It is the wider definition of well beings, which is either based on the material deprivations or something which can be publicly provided.
In the view of Alvan (2009), human development is a process of enlarging people choices and increasing their freedom of choices.It can be done by expanding human capabilities and contestability.However, there is a need to gauge achievement with respect to capabilities and contestability.In this connection, the concept of Human development index is developed in the literature.It is used to measure average achievement in the country with respect to three dimensions.These includes long and healthy life, knowledge and a decent standard of livings, (Grimm et al., 2008).
There is a vast body of literature on the studies connecting human development to a number of variables such as economic growth, poverty, income inequality and so on.The studies are diverse in terms of methodology used, data type, sets of variables and area of the study etc.However, results of these studies can be compared up to some extent.For instance, Novid and Sumarsono (2018) established a positive effect of Human Development Index (HDI) on economic growth in the East Java Indonesia.Other studies which reported the similar results includes the study of Ranis (2004), who reported that relationship between HDI and economic growth is the two ways.There can be either high growth and large gains in human development (HD) -known as virtuous cycle or low growth and low gain in HD which is known as vicious cycle of development.
In a study of 35 to 76 developing countries for the period of 1960-92, Ranis et al., (2000) concluded that economy can be a mutually reinforcing upward spiral of high level of HD leading to high growth and high growth further promote HD.Suri, Boozer, Ranis, and Stewart (2011) also supported Ranis et al., (2000) view that initial investment in HD can lead to accelerating economic growth, and economic growth in turn improve status of HD of a country.
In the relationship from HD to economic growth, Zhang (2019) reported that there is a positive impact of HD on economic growth in the period of 1990-16 in the 29 developing Asian countries.In different methodology, Iskandar (2017) reinforced Ranis et al., (2000) findings in the study covering 23 district/cities of Aceh province of Indonesia.Meanwhile, Bansal et al., (2021) reported long run positive relationship between human development and economic growth for the South Asian countries for the period of 1990 to 2007.
Human development (HD) can also, via channel of affecting GDP, effect income inequality, which in turn, may influence level of HD in a country.The two-way relationship between HD and income inequality is also reported by Alvan (2009).In the panel data analysis of ninety countries, Alvan (2009) reported that HD and income inequality are negatively related and the causality runs from both the direction.Faisal (2022) reported that HD positively affect G in the sample of eight countries of SAARC group.
In this paper, the authors make an attempt to find out the relationship among the human development, income inequality and economic growth for the SAAR organization along with China for the period of 1990 to 2020.The inclusion of China, sample of countries, is to balance the effect of Indian economy as it is the largest economy in the SAARC member States.The main objective of the study is to find out the role of human development in promoting economic growth and reducing income inequality in the said sample.As per the knowledge of the authors, there is no such paper which address analyze the issue in the present settings, it is, therefore, contribution to the body of already available literature.

Methodology
The present study investigates the relationship between human development index (HDI), Income Inequality (G) and Economic growth (EG) in the SAARC countries.The following regression equations are developed on the basis of literature review.For instance, equation ( 1) and ( 2 Similarly, equation ( 3) is derived from Roine, Vlachos and Waldenström (2009) as they investigated the determinants of income inequality of 16 countries over the entire twentieth century.Their model is modified by including human development index as an explanatory variable along with dynamic panel specification at level and various lags of variables.Equation ( 1) to (3) are not treated as simultaneous equations rather estimated as single equation models as did by Binder and Georgiadis (2010).The equations contain lag terms of various variables which can reduce the issue of simultaneity in the regression.
The variables HD, G and EG represent human development index, income inequality measured by Gini Coefficient and economic growth of each country.Economic growth is measured by log of Gross National Income per capita based on PPP of 2007.Whereas, X is the vector of explanatory variables.In equation (1), X includes mean years of schooling (MYS) and Gender Development Index (GDI).In equation (2), X includes population growth rate (in percentage) measured by P and control of corruption (denoted by CC).In equation (3) includes agricultural land percentage of total area denoted by AGL.The panel data set consist of SAARC bloc along with China.The country china is included to balance the economic size of India in the sample as India one of the largest economies in the SAARC bloc.
The data on HD, G, EG, GDI, MYS are obtained from Human Development Reports (various issues) by UNDP.In case of some variables and countries data was not available for 2019 and 2020, then the average of the last five years data was used as a proxy.However, data on P and AGL is taken from World Development Indicators (various issues) by the World Bank.Data on CC is obtained from governance indicators, published by the world 10.57030/23364890.cemj.31.2.19 156 | P a g e bank.It is to be mentioned that data was not available from 1990 to 1996, and average of the five years was taken as a proxy for these years.The time period of analysis is from 1990 to 2020 with total 9 countries in the panel.

Discussion
The analysis starts with testing for the existence of unit root test in the time series.In this context, Levin, Lin & Chu ( 2002) is used as it more powerful test in the moderate size of panel data.The results of the test are given in the Table No.1.Test has been performed on three options, namely, intercept, both intercept and trend and None.These results are reported in Table No.1.It is evident from the table that most of the series are stationary at level with intercept term included in the regression of the test statistic.However, variable "P" is stationary at level when both intercept and trend terms are not included in the regression equations.The variables "MYS" is stationary at level when intercept and trend term is placed in the equation of test statistic.The overall conclusion is that all the series are stationary at level and do not suffer from the unit root issues.It is, therefore, no need to test for cointegration among of the variables.In the next phase of estimation, the testing for Fixed Effect (FE) versus Random Effect (RE) are vital to undertake.In this connection, the Hausman (1978) test has been used.The results are of those models are reported which specifications are found reliable on the basis of Hausman (1978) test.In Table 2, the estimated results of the equation (1) are reported.There are four specifications of the equation (1) which are fixed effect as confirmed by Hausman (1978) test.One of the benefits of estimating different specifications of the model is to see whether sign, size and significance of the model changes with changes in the specifications are not.It can be seen from the Table 2 that constant term is negative in all the four specifications, and highly statistically significant except one specification.Moreover, in three case, the size of the coefficient is very similar.
The impact of income inequality (G), measured by Gini coefficient is negative on human development index (HD) in the sample of nine countries is negative and highly statistically significant in three specifications.In the literature, the same is reported by Tripathi (2021), Castells-Quintana, Royuela & Thiel (2019), Faisal (2022) and Saragih (2018).It means that when value of the Gini coefficient approaches to 1, the distribution of income in the society becomes pro-rich.In this context, as income inequality rises, HD falls.One of the reasons is the inability of poor to spend on education and health.The results obtained are in accordance with the earlier studies.
The impact of economic growth (EG) on the HD is positive and significant except one specification of the model.It means that when economic growth accelerates, HD score or rank of the country also improves.In one of the specifications, one-year lag value of the EG is also included.Its sign is positive but it is statistically not significant.the results are in accordance to Bansal et al., (2021), Ranis et al., (2000), Suri, Boozer, Ranis, Stewart (2011), Zhang (2019) and Iskandar (2017).
GDI measure inequalities in achievement in three basic dimensions of human development.The effects of GDI at time "t" and "t-1" on HD are positive and significant.These finding is also reported by Bhowmik (2020) for the sample of 12 developing countries for the period of 1990 to 2015.
In Table 2, the sign of MYS at time "t" is positive and significant, however, it is negative and significant at time "t-1".In other words, the coefficients on the current value of MYS are positive and statistically significant, and coefficients on one-year lag values of MYS are negative and statistically significant.In the literature, Jalil and Kamaruddin (2018) reported that mean years of schooling has positive and significant effect on human development in the case of fifteen selected developing countries for the period of 2010 to 2014.Table 3 contains the results of equation ( 2), in which dependent variable is EG.The explanatory variables are G, HD, P and CC and their lag values as explanatory variables.There are three specifications of the equation (2).Before estimating these models, Hausman (1978) test was conducted to know whether FE or RE model is appropriate.It is indicated, by the Hausman (1978) test, that FE model is appropriate.The value(s) of Hausman (1978) test(s) is(are) not reported here.
In the Table 3, it is evident that the sign of coefficient "G" is positive but not statistically significant.In case of HD, the sign of coefficients on the current values of HD are all positive and significant.It means that improvement in the index of HD accelerate economic growth in the panel of 9 countries.These results are as per the studies of Ranis et al., (2000) Iskandar (2017).In economic theory, it is argued that human development can cause economic growth to accelerate in the first phase, and later on, economic growth, through expansion in the level of income of households and government, increase spending on the goods related to expanding human development and capabilities.
In case of population growth P, the study indicates that its effect is negative and significant (see Table 3).It means that rapidly growing population can reduce economic growth in the panel of the 9 countries.The study of literature, as reported by Headey and Hodge (2009), indicate that effect of population growth on economic growth is not robust.It can be positive or negative, however, it is significantly different than zero.
Corruption is considered to be one of the determinants which can slow down economic activities of a country, (Mo, 2001).Gründler and Potrafke (2019) employed a new data set for 175 countries for the period of 2012-2018 and concluded that corruption negatively affect economic growth.But the present study does not report any evidence in the support of the theory.However, in the present case, the coefficients of control of corruption, at time t, are all positive but statistically insignificant.Moreover, the coefficient of control of corruption, at time t-1, is negative but statistically insignificant.The determinants of G are given in the Table 4.These are EG, HD and AGL and their lag values.The link between economic growth and income inequality is largely debated in the literature.These views are divided into positive and negative effect of economic growth on income inequality in different types of data sets, research methodology and period of analysis.For instance, Deininger and Squire (1997) concluded that there is no evidence that economic growth has negative effect on income inequality.Moreover, they also reported that equal access of people to credit market and their ability to invest can promote economic growth.Finally, they found that redistribution policies aimed at helping poor can accelerate economic growth if it does not hamper investment.In the present case, as evident from the Table 4, effect of EG, at time t, on G is positive and significant.It means that expansion in the economic activities are causing income inequality to rise in the panel of 9 countries.However, the effect of EG, at time t-1, on G is negative and significant in one case.
The impact of HD on G is negative.All the coefficients are statistically significant except only in one specification.It means that improvement in the human development index causes income inequality of panel of 9 countries to fall.In other words, income inequality can be reduced if efforts are made to improve ranking on the human development.This is as per the study of Faisal (2022), The impact of AGL on G is also given in the Table 4.It is clear that coefficients of AGL are negative and statistically significant.AGL is the agricultural land as percentage of total land.If proportion of agriculture land in total land increases, the income of the rural households associated with agriculture also increases.It can improve income distribution in the favor of marginalized class.10.57030/23364890.cemj.31.2.19 159 | P a g e The panel data analysis of relationship among human development, income inequality and economic growth shows that income inequality significantly and negatively affect human development, whereas, economic growth, gender development index and mean year of schooling positively and significantly affect human development in the sample of the countries.In the equation of determinants of economic growth, human development positively and significantly contributes to the economic growth of the sample countries.In the equation of income inequality as a dependent variable, economic growth is the significant cause of expanding income inequalities, whereas, human development and agriculture area as percentage of total area is significantly reducing income inequality in the countries.
) are replicated from Binder and Georgiadis (2010).They estimated determinants of human development for 84 countries from 1970 to 2005.They used auto-regressive model, whereas, the present paper uses dynamic panel version of the Binder and Georgiadis (2010).