Are Human Development and Innovativeness Levels Good Predictors of the Competitiveness of Nations? A Panel Data Approach

Abstract

Nations must adapt to the changing and developing world to sustain and develop their competitiveness. Human development and innovation are the two key concepts to increase the competitiveness of a nation. This study aims to examine the relationship between the Human Development Index (HDI), Global Innovation Index (GII), and Global Competitiveness Index (GCI) across different income groups from 2010 to 2019. The main objective is to identify potential variations in these relationships based on the income level of the countries involved. Panel data analyses using Common Correlated Effects Mean Group (CCEMG) and Augmented Mean Group (AMG) estimators are conducted to examine the relationships. Additionally, Pairwise Dumitrescu Hurlin Panel Causality Tests are conducted to examine the causal relationships between variables. The results show that HDI has a significant positive effect on GCI in each income group. Improving human development such as raising living standards and providing equal education opportunities for every member of society can contribute to a country’s competitiveness. Moreover, it is found that the effect of GII on GCI varies by income group. Specifically, the results indicate that the effect of GII on GCI is not supported for upper-middle-income countries. Therefore, while developing strategies to increase competitiveness through innovation, it is important to consider the income group of a nation. The findings of this study may assist policymakers, researchers, academics, and politicians to enhance their perspectives and formulate strategic and effective recommendations for action.

1. Introduction

Nations are challenged by the changing and developing world to maintain and increase their level of prosperity. Furthermore, nations pursue protection of what they have, expediently meeting basic needs, and seeking the greatest future since ancient times. Some characteristics such as geographic location, country size, and natural resource size that affect prosperity are inherited [1] while some properties are created by efforts and investments. Moreover, some countries such as Singapore, Israel, Finland, Sweden, and certain Gulf nations, although lacking inherent advantages like natural resource abundance or strategic geographic positioning, excel in national prosperity. A cross-country analysis by Sachs and Warner [2] reinforces the idea that countries rich in natural resources, as indicated by a high ratio of natural resource exports to GDP, tend to experience slower economic growth compared to resource-poor counterparts. An alternative viewpoint regards natural resource endowment as a potential advantage, exemplified by countries like Botswana, Norway, Chile [3], the USA, Canada, and Australia [4], which are acknowledged for effectively utilizing their natural resource wealth to foster robust growth [5]. The aforementioned dual viewpoint emphasizes the significance of efficiently utilizing and diversifying resources to achieve sustainable development. In addition, even a vast amount of inherited wealth can vanish while keeping up with globalization. For this reason, it has become a necessity for nations to compete not only with other nations but also within themselves to race in this global contest. Competitiveness has been thriving progressively [6]. After Porter’s [7] definition of a national level of competitiveness, the concept became popular [8], and due to persistent usage of this term by the media, politicians, and businesspeople [9], a need for research has emerged in various aspects [6]. However, there is no consensus on the definition [8,9] because of the broad scope and vagueness of the term national competitiveness [9], although it has been discussed for years by scholars. The definition of national competitiveness has evolved in conjunction with the determinants of competitiveness [10]. Porter’s [7] definition of “the only meaningful definition of competitiveness at the national level is national productivity” has guided a vast amount of research. Later, Krugman [11] criticized the idea of the productivity of nations and claimed that firms compete for market share, not nations. Additionally, Kohler [12] claimed that countries are not firms written in large letters, and he criticized using competitiveness as equal to domestic welfare or productivity as a misleading paradigm and stated that achieving higher levels of domestic welfare or productivity did not mean increasing international competitiveness. Moreover, Aiginger [13] contributed to the definition of national competitiveness by using soft factors including capabilities, quality of institutions, national innovation system, trust, political stability, the rule of law, and hard factors concerning production function regarding capital, labor force, and technical progress. Besides, there have been other studies emphasizing not only economic indicators, such as Janger et al.’s [14] work, but they also defined competitiveness as the “ability to raise standards of living and employment, while maintaining a sustainable environment and sustainable external balances”. Meanwhile, Delgado et al. [1] introduced a new framework combining microeconomic competitiveness and macroeconomic competitiveness indicators and defined foundational national competitiveness as “the expected level of output per working-age individual given the overall quality of a country as a place to do business”. Additionally, Aiginger and Vogel [9], as a result of their comprehensive research, proposed a new framework including price competitiveness (cost and productivity), quality competitiveness (structure and capabilities), and outcome competitiveness (traditional and new perspectives) under the idea of different steps of the evaluation process of competitiveness. Hence, there has been increasing acknowledgment that only price-related concepts alone are not a sufficient measure of a country’s competitiveness.

As a result, the concept of competitiveness is now often used to refer to a more comprehensive and multi-faceted concept that goes beyond just economic productivity. Therefore, describing national competitiveness without including human force, technology developments, government policies, and infrastructure of the nations is similar to a puzzle with missing pieces. Consequently, in this contemporary environment, a country’s competitiveness is largely dependent on its technological advancements [15] and human resources [16]. Being essential for the welfare of national economies [17], innovation is an important driver of growth and competitiveness [18] as well as a vital tool for gaining international competitive advantages and sustainable economic growth [19]. Furthermore, investing in human development is another key factor besides innovation to acquire competitiveness. As Fagerberg [20] mentioned, securing a higher living standard for citizens for the present and future is a capability of nations. As mentioned above, competitiveness, human development, and innovation are linked, and investigating these concepts while eliminating their relationships is inadequate.

In addition, in the contemporary global economic environment, the importance of examining national competitiveness by income levels has emerged. Given problems such as the large gap between the growing population in developing countries and the aging population in developed economies, the inability of developing countries to keep up with the pace of digitalization in developed economies and the widening gap between them, the worsening of significant talent gaps in both developed and developing countries, but not the same ability to import talent has been underlined by the World Economic Forum (WEF) [21]. To gain a deeper understanding of the significance of competitiveness in relation to national income levels, nations are categorized based on the income categorization provided by the World Bank. The objective of categorizing nations based on their economic levels is to discern the distinct impacts of human development and innovation on their competitiveness. Using a variety of analyses, the literature presents several examples comparing income groups. These include investigations into the correlation between GDP and health expenditure among a sample of 74 countries [22], examinations of the socioeconomic factors influencing carbon emissions in 64 countries [23], and explorations of the relationship between Innovation Input and Output of countries using diverse datasets from different years [24].

For this reason, in this study, the effects of human development and innovation on competitiveness at the national level between 2010 and 2019 including 121 countries based on their income levels with a panel data approach are examined by utilizing Human Development Index (HDI), Global Innovation Index (GII) and Global Competitiveness Index (GCI).

The contributions of this study to the empirical literature are thus threefold. First, we analyze the effect of HDI and GII on GCI including 121 countries between 2010 and 2019. Second, more sophisticated econometric techniques such as Common Correlated Effects Mean Group (CCEMG) [25,26] and the Augmented Mean Group (AMG) estimators [27] are utilized, and comparisons of the results are drawn. Third, countries are divided into groups based on their income levels, and to our knowledge, this is the first study to examine and compare the relationships among HDI, GII, and GCI based on income levels. Thus, in contrast to similar studies, this research aims to provide a distinctive contribution to the literature by taking into account the income levels of nations. By utilizing this perspective, the study seeks to offer new insights to countries, suggesting that strategies for enhancing competitiveness may vary depending on a nation’s economic dynamics. The aim is to inspire novel strategies for countries at different levels of economic development, consequently bridging the gap between research and practical insights.

This paper is organized as follows: Section 2 introduces the theoretical background of the subject and related national indices with examples from the literature; Section 3 presents the research objective, methodology, and data; the following section discusses the obtained results; and finally, in the last section, conclusion and recommendations are mentioned.

2. Theoretical Background

There have been dramatic changes in the approaches used to define national competitiveness over time. While it was previously considered only as economic growth, it has become a concept in which macro- and micro-economic indicators are used together in the meantime. Ultimately, national competitiveness has evolved into a comprehensive concept including various aspects (e.g., productivity, institutions, competencies, technological readiness, infrastructure, education, and training) with one context. Even reaching consensus on the definition of national competitiveness has been a long process; the measuring of national competitiveness has been in progress regarding expectations of the era. For many years, ranking the competitiveness of nations and the factors affecting their competition has been in the spotlight [28]. Globalization, fast technical change, and diminishing economic distance trigger policymakers to express serious concerns about national competitiveness [29]. A tool for systematic ranking for nations to monitor their competition and evaluate their relative success based on pre-determined criteria has been needed. Diverse stakeholders have utilized ranking indices for different purposes such as determining investment plans by the business community, benchmarking policies to attract enterprises by governments, and employing cross-country analyses by academicians [29]. Two widely accepted and more influential indices have been prepared by the World Economic Forum (WEF) and the International Institute for Management Development (IMD) for years [28,30,31]. Also, there are very few similar indices that are found in the literature; however, they are not able to sustain their continuity such as the Centre for International Competitiveness [32] or not finding an application area by other researchers [31,33]. Besides these indices, there are many developed by governments, research institutions, and consultants to serve as a guideline for strategy development [29]. Therefore, the Global Competitiveness Index (GCI) published by WEF and the World Competitiveness Year Book (WCYB) of the IMD are the main dominators of the field [30]. Both indices have measured competitiveness with qualitative and quantitative indicators including enormous data. Despite many decades of existence in the field, neither GCI nor WCYB became out of date due to their agile and evolving indicators; they updated their indices based on the requirements of the era.

Since 1979, the WEF has published a Global Competitiveness Report annually to provide an assessment regarding the strengths and weaknesses of nations in terms of competitiveness [28]. Annual publishments transformed into the Growth Competitiveness Index, Business Competitiveness Index, and finally evolved to the Global Competition Index including microeconomic and macroeconomic factors that are supported by twelve basic pillars that are categorized into three sub-indices called Basic Requirements, Efficiency Enhancers, and Innovation & Sophistication. These pillars include not only quantified data (hard data) supported by internationally recognized agencies such as the International Monetary Fund (IMF), United Nations Education, Scientific and Cultural Organization (UNESCO), and the World Health Organizations, but also qualified data (soft data) composed by the executives of the analyzed nations [31]. Moreover, WCYB has been published annually since 1989 to serve as a reference point for benchmarking nations by using statistics and survey data, and IMD World Competitiveness Ranking (WCR) is structured on four main sub-factors of Economic Performance, Government Efficiency, Business Efficiency, and Infrastructure to measure competitiveness with more than 330 criteria using hard data, surveys, and background information [34]. In this research, GCI is utilized for national competitiveness measurements due to respectful academic press, doyen academicians behind the brain team, and openness of information. WEF shares reports, methodology, and data more openly and in detail for stakeholders relative to IMD.

GCI assists as a tool for a nation to monitor its relative place among nations as well as its progress through time. There are no losers in this competition; even understanding the current position is a first step to being in this race for a nation. Accordingly developing strategies and taking actions are pushful movements followed by nations. Premise actions should also be considered by assessing the factors driving competitiveness.

In recent years, the role of innovation has been pointed out to be one of the main drivers of global competitiveness in the literature and the practical field [35]. As Dobrovic et al. [36] indicated, a great number of authors emphasized that innovation plays an important role in competitiveness. Innovative countries tend to have more productive and efficient economies, which in turn improves their competitiveness. Therefore, a country’s performance in the GII can be used as an indicator of its potential competitiveness in the global market. Similar to GCI, the Global Innovation Index (GII) was created to measure several components of the innovation ecosystem among nations in cooperation with INSEAD, the World Intellectual Property Organization, and the SC Johnson College of Business at Cornell University [37]. Furthermore, GII is widely accepted as the data source to examine national innovation competence [38]. Also, with an increasing number of countries recognizing innovation as an important driver of their economic growth, the GII is expected to continue to grow in importance in the coming years [39]. The Global Innovation Index (GII) is composed of two sub-indices, which are the Innovation Input and Innovation Output sub-indices. The Innovation Input sub-index measures factors that facilitate innovative activities in a country, while the Innovation Output sub-index represents the outcomes of those innovative activities. The Innovation Input sub-index comprises five elements: Institutions (Is), Human capital and research (HCR), Infrastructure (Ir), Market sophistication (MS), and Business sophistication (BS). On the other hand, the Innovation Output sub-index has two elements: Knowledge and technology outputs (KTO) and Creative outputs (CO). Each of these elements consists of several sub-variables [37].

Moreover, one of the other indicators to increase national competitiveness is investing in human development. However, there has been a limited number of papers on the relationship between competitiveness and human development in the existing literature [40]. Notionally, human development has a positive impact on global competitiveness since qualified and educated people enhance global competitiveness by creating values [41]. Therefore, people and their valuable contributions are the keystones of national competitiveness [42]. The presence of well-educated and healthy human capital is a crucial element in the formation of competitive advantages for a nation [43]. Specifically in developing countries, lack of human development acts as a barrier to nations’ technology catchups [44]; therefore, in a globalized world, this means starting competition without one of the most important tools necessary. Additionally, since human development is known to be associated with economic growth, increasing human development is a priority in national and international development strategies in developing countries to achieve the aspirations of the 2030 Agenda for Sustainable Development [45].

In the same manner as GCI and GII, there is an index developed for measuring human development. The Human Development Index (HDI) is the most used and globally recognized index for measuring social development in society [46]. Since 1990, it has been a comprehensive measure of human development that takes into account three fundamental dimensions: access to knowledge (education), a long and healthy life (health), and a decent standard of living (income) published by United Nations Development Programme (UNDP) [47]. The coverage of the HDI is determined by data availability, and to enable cross-country comparisons, the index is calculated based on data from credible sources, including leading international data agencies.

In the literature, there are several studies focusing on analyzing the relationships among HDI, GII, and GCI employing several methods. Nasierowski [48] examined several composite indexes including HDI, GII, and GCI to investigate their relationships by employing Pearson correlations and found that there is a strong relationship between the HDI and GII, the HDI and GCI, and the GII and GCI. Additionally, Fonseca and Lima [49] found a high positive correlation between the GII and the GCI. Similarly, Onyusheva [50] conducted a Pearson correlation analysis to capture the relationships among these indices between 2008 and 2016 for Kazakhstan and found a strong relationship between these indices. Kiselakova et al. [51] studied the relationship of GCI to several indices along with GII with European Union-28 economies throughout 2014–2018 and discovered that GCI was mainly influenced by GCI. Also, few studies have examined causal relationships among these three indices. Taranenko [52] utilized a regression model to examine the effect of GII indicators on GCI by using data from 2012 and found that general infrastructure, creative intangibles, and investment had a positive impact on GCI. Furthermore, Cvetanovic et al. [53] examined the relationship between GII and GCI of six Western Balkan countries along with a group of six selected European Union (EU) countries in 2012 by employing regression analysis and discovered that GII has a positive but weak effect on GCI. Additionally, Cetinguc et al. [54] examined the relationships among HDI, GII, and GCI by using the PLS-SEM method for data from 2015 and found that HDI has a positive impact on GII and GCI whereas GII has a significant effect on GCI.

Moreover, there are examples of panel data analysis on the relationship of the beforementioned indices: Kiselakova et al. [16] examined the effects of HDI, doing business index (DBI) on GCI of 28 countries in the European Union with panel data analysis for the period of 2006–2017, and observed that HDI positively affected GCI. Further, Hamid [41] examined the effects of HDI and its factors on GCI via a panel data approach for ten ASEAN countries for 2010–2015 and found that HDI positively affected GCI.

3. Research Objective, Data and Methodology

The main objective of this research is to investigate and compare the effects of human development and innovation on competitiveness at the national level between 2010 and 2019 including 121 countries in different income groups. We select this period for two reasons: to avoid unpublished years of indices before 2010 and to refrain from the impact of the COVID-19 pandemic.

In this study, we use a 10-year balanced panel data set covering 121 countries from 2010 to 2019. In addition, we split countries into three groups according to their level of income based on the World Bank classification. This classification is updated each year based on the gross national income (GNI) and annual inflation rate (using the SDR deflector). GNI measures are expressed in US dollars (USD) and are determined using conversion factors derived according to the Atlas method. In this study, the classification of 2019 is applied to categorized countries. A total of 217 countries are included in the 2019 country classification list. In this classification, countries with a per capita GNI below $1025 in 2018, calculated using the World Bank Atlas method, are low-income countries, countries with that between $1026 and $3395 are lower-middle-income, countries with that between $3396 and $12,375 are upper-middle-income countries and, finally, countries with a per capita GDP of more than $12,375 are also defined as high-income economies. Additionally, the data related to competitiveness are collected from the GCI, data related to innovativeness are derived from the GII, and finally, human development measures data are gathered from HDI. Subsequently, all available data are aggregated together and countries without data of more than one index are eliminated. If one country had a few years of missing data in one particular index, that missing data were replaced by averages of available data. After the data-cleaning process, a total of 121 countries remained.

Table 1. Distribution of the Studied Countries by Their Income Level.

High Income		Upper-Middle Income		Lower-Middle Income
Australia	Latvia	Albania	Mexico	Angola	Tunisia
Austria	Lithuania	Algeria	Montenegro	Bangladesh	Ukraine
Bahrain	Luxembourg	Argentina	Namibia	Bhutan	Vietnam
Barbados	Malta	Armenia	Paraguay	Bolivia	Zambia
Belgium	Netherlands	Azerbaijan	Peru	Cabo Verde	Zimbabwe
Brunei	New Zealand	Bosnia and Herzegovina	Romania	Cambodia
Canada	Norway	Botswana	Russia	Cameroon
Chile	Oman	Brazil	Serbia	Côte d’Ivoire
Croatia	Panama	Bulgaria	South Africa	Egypt
Cyprus	Poland	China	Sri Lanka	El Salvador
Czech Republic	Portugal	Colombia	Thailand	Ghana
Denmark	Qatar	Costa Rica	Turkey	Honduras
Estonia	Saudi Arabia	Dominican Republic	Venezuela	India
Finland	Seychelles	Ecuador		Indonesia
France	Singapore	Gabon		Kenya
Germany	Slovak Republic	Georgia		Kyrgyz Republic
Greece	Slovenia	Guatemala		Lesotho
Hong Kong SAR, China	Spain	Guyana		Moldova
Hungary	Sweden	Iran		Mongolia
Iceland	Switzerland	Jamaica		Morocco
Ireland	Trinidad and Tobago	Jordan		Myanmar
Israel	United Arab Emirates	Kazakhstan		Nicaragua
Italy	United Kingdom	Lebanon		Nigeria
Japan	United States	North Macedonia		Pakistan
Korea	Uruguay	Malaysia		Philippines
Kuwait		Mauritius		Senegal

lists countries based on their income levels.

In the data preparation process, the use of different scales, both within the indices and relative to each other, comes to prominence as one of the most important issues to be considered. For instance, GCI employs a 1–7 scale between 2010 and 2017 and later switched to a 0–100 scale in 2018 and 2019. Moreover, GII utilizes a 1–7 scale only in 2010, and the rest of the years are measured by a 0–100 scale. On the other hand, HDI applies a 0–1 scale for each year. To be gathered under the same scale, all scales are normalized to a 0–1 scale. Descriptive statistics of the main variables are given in

Table 2. Descriptive Statistics. Source: Own Research.

	All Countries		High-Income		Upper-Middle-Income		Lower-Middle-Income
	Mean	Std. Dev.	Mean	Std. Dev.	Mean	Std. Dev.	Mean	Std. Dev.
GCI	0.571835669	0.107047548	0.65681378	0.089461961	0.537334927	0.063790498	0.47543713	0.06152495
HDI	0.767541322	0.119264365	0.875015686	0.046987769	0.749289744	0.04778229	0.613690323	0.076879435
GII	0.384162001	0.112123124	0.482448862	0.093970335	0.334549269	0.05260266	0.284880282	0.047105741

. There is a descending order of mean GCI scores from high-income (Mean = 0.6568) to upper-middle-income (Mean = 0.5373) and lower-middle-income countries (Mean = 0.4754), indicating higher competitiveness in wealthier nations. High-income countries exhibit a relatively lower standard deviation (Std. Dev. = 0.0895), implying a more concentrated distribution of GCI scores and potential homogeneity in economic competitiveness. In contrast, the larger standard deviations in upper-middle and lower-middle-income categories (Std. Dev. = 0.0638 and 0.0615, respectively) suggest a more diverse economic landscape within these income strata. Upon examination of the HDI, a clear stratification is observed, with high-income countries displaying the highest average (Mean = 0.8750), followed by upper-middle-income countries (Mean = 0.7493) and lower-middle-income countries (Mean = 0.6137). High-income countries (Std. Dev. = 0.0470) and upper-middle-income countries (Std. Dev. = 0.0478) exhibit a relatively lower standard deviation, indicating a more homogeneous distribution of human development achievements. In contrast, lower-middle-income countries (Std. Dev. = 0.0769) exhibit larger standard deviations, meaning that human development values in this group of countries vary more. Similar to HDI, a discernible hierarchy emerges, with high-income countries exhibiting the highest mean GII (Mean = 0.4824), followed by upper-middle-income nations (Mean = 0.3345) and lower-middle-income countries (Mean = 0.2849). In contrast to other indices, the standard deviations seen in the GII demonstrate a decreasing trend as one transitions from high-income nations to lower-income nations. This indicates that high-income countries exhibit a more pronounced degree of divergence in terms of innovation.

By drawing upon the related literature mentioned above, we adopt the following econometric model in Equations (1) and (2):

$${GCI}_{it}={\alpha }_i+{\beta }_{yi}{GII}_{it}+{\beta }_{zi}{HDI}_{it}+u_{it},$$

(1)

$$u_{it}={\mu }_i+{\phi }_i{\vartheta }_t+{\epsilon }_{it}.$$

(2)

• ${GCI}_{it}$: GCI normalized scores of country i in year t,
  ${GII}_{it}$: GII normalized scores of country i in year t,
  ${HDI}_{it}$: HDI normalized scores of country i in year t,
  ${\alpha }_{it}$: intercept for country i,
  ${\beta }_{yi}$: Slope coefficient on GII,
  ${\beta }_{zi}$: Slope coefficient on HDI,
  $u_{it}$: error component term,
  ${\mu }_i$: unobserved individual effect,
  ${\phi }_i$: factor loadings,
  ${\vartheta }_t$: unobserved common effects,
  ${\epsilon }_{it}$: represents the error term.

Although there has been adequate research regarding competitiveness, innovation, and human development, mainly cross-section analysis has been preferred by researchers [48,52,53]; there have been very few studies using the indices as stated in this study with panel data approach [41,51,55]. Consequently, panel data analysis has recently been used in the relationships on various indices in the literature. Due to the advantages of panel data, such as over cross-section data, being a more precise inference of model parameters and greater capacity for capturing the complexity of behaviors if data are available [56] are the reasons for being selected, also rare usage of these models related to our research are the main motivators for us to choose panel data techniques to utilize in this research.

In estimating and testing with panel data models, cross-section dependence has great importance. O’Connell [57] showed that disregarding cross-sectional dependence can cause over-rejection of the null hypothesis of a unit root in panel data. To consider this issue and avoid the over-rejection problem, we test the absence of cross-section dependence across the different income-level groups, applying the cross-section dependence test developed by Pesaran [58]. We use this test since when cross-sections are greater than time (N > T), Pesaran’s test [58] is more appropriate to use. Then, we utilize the CIPS test (second-generation panel unit root test) developed by Pesaran [59] to test the non-stationarity of the variables, as this test allows for cross-section dependence across the different income level groups. Additionally, slope heterogeneity is another issue to consider. In our model, we assume slope heterogeneity to allow coefficients of variables to vary across each country. Thus, we employ the slope homogeneity test proposed by Pesaran and Yamagata [60] as well as Blomquist and Westerlund [61] which also considers heteroskedasticity and autocorrelation to test the absence of heterogeneous slope coefficients.

To explore the impact of human development and innovation on competitiveness, we use the Common Correlated Effects Mean Group (CCEMG) estimator developed by Pesaran [25] and the Augmented Mean Group (AMG) developed by Eberhardt and Bond [62] and the Eberhardt and Teal [27,63] estimator. The reasons for selecting these estimators are that they allow both slope heterogeneity and cross-section dependence across all individuals. In addition, we utilize the panel causality test advanced by Dumitrescu and Hurlin [64] to reveal the pairwise causal relation between human development, innovation, and competitiveness. Finally, we employ the Two-Way Fixed Effect Method [65] to check the robustness of the findings of AMG and CCEMG estimations.

4. Results

The pairwise correlations of the main variables are shown in

Table 3. Correlations of Main Variables.

	All Countries			High-Income			Upper-Middle-Income			Lower-Middle-Income
	GCI	HDI	GII	GCI	HDI	GII	GCI	HDI	GII	GCI	HDI	GII
GCI	1			1			1			1
HDI	0.8138 ***	1		0.7925 ***	1		0.3790 ***	1		0.6230 ***	1
GII	0.8519 ***	0.8328 ***	1	0.7313 ***	0.8659 ***	1	0.6108 ***	0.3623 ***	1	0.6037 ***	0.6148 ***	1

*** Indicates 1% level of significance.

. Correlations among the main variables are higher for all countries and high-income countries, in contrast to those of upper-middle-income and lower-middle-income countries. The observed higher correlations among the main variables underscore the nature of the relationships within the dataset. The proposed approach in this study permits common effects to vary in their impacts on individual units, simultaneously allowing them exhibition of arbitrary degrees of correlation with both themselves and individual specific regressors; furthermore, it accommodates serially correlated and heteroscedastic errors at the individual level, without imposing the requirement of identical and/or independent distribution of individual specific regressors across cross-section units [25].

Pesaran [58] test was used to test weak cross-section dependence along with Pesaran [66,67] test was used to test cross-section dependence.

For all variables’ absence of cross-section dependence is strongly rejected as seen in

Table 4. Cross-Section Dependence. Source: Own Research.

CSD	All Countries		High-Income		Upper-Middle-Income		Lower-Middle-Income
	Pesaran [58]	Pesaran [66,67]	Pesaran [58]	Pesaran [66,67]	Pesaran [58]	Pesaran [66,67]	Pesaran [58]	Pesaran [66,67]
GCI	233.896 ***	187.98 ***	101.618 ***	88.39 ***	75.982 ***	60.88 ***	56.647 ***	37.36 ***
GII	202.130 ***	222.25 ***	85.254 ***	97.23 ***	67.669 ***	60.24 ***	50.652 ***	63.08 ***
HDI	254.452 ***	25.37 ***	107.833 ***	15.48 ***	78.189 ***	17.21 ***	66.436 ***	5.07 ***

*** Indicates 1% level of significance.

. Cross-sectional dependence is prominent across all income groups for all variables with high significance at a significance level of 1%. Therefore, we take into account cross-section dependence among all countries—high-income, upper-middle-income, and lower-middle-income groups.

After verifying cross-section dependence, the panel unit root test proposed by Pesaran [49] was employed to determine whether the variables were stationary or not.

Table 5. Results of the Unit Root Test CIPS. Source: Own Research.

	All Countries			High-Income			Upper-Middle-Income			Lower-Middle-Income
CIPS	GCI	GII	HDI	GCI	GII	HDI	GCI	GII	HDI	GCI	GII	HDI
t_stat	−287.7	−329.7	−252.1	−269.9	−271.6	−341.0	−291.8	−301.0	−299.8	−402.0	−288.1	−298.0
Critical value at
10%	−2.03	−2.03	−2.03	−2.07	−2.07	−2.07	−2.10	−2.10	−2.10	−2.12	−2.12	−2.12
5%	−2.13	−2.13	−2.13	−2.19	−2.19	−2.19	−2.22	−2.22	−2.22	−2.25	−2.25	−2.25
1%	−2.32	−2.32	−2.32	−2.41	−2.41	−2.41	−2.47	−2.47	−2.47	−2.51	−2.51	−2.51
p-value	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01	<0.01

illustrates the results of unit root tests under the null hypothesis that CIPS values are nonstationary. The results show that all variables for each income group are stationary.

In panel data econometrics, it is important to consider the issue of slope heterogeneity, which occurs when the parameters of interest vary across cross-sectional units due to different economic and demographic structures. When using large panel data sets with a small number of time periods (T), it is not valid to assume that the slope parameters are homogenous. If this assumption is not tested and the data do contain slope heterogeneity, the estimates may be misleading [68]. To test the homogeneity of each slope coefficient among the income groups in our model, the slope homogeneity test introduced by Pesaran and Yamagata [69] and Blomquist and Westerlund [61] was utilized. Under the null hypothesis of the slope, coefficients are homogenous. Slope homogeneity results are given in

Table 6. Slope Homogeneity. Source: Own Research.

	All Income	High-Income	Upper-Middle-Income	Lower-Middle-Income
Delta_tilde	−10.882 ***	0.043	1.666 *	1.827 *
Adjusted_Delta_tilde	−14.049 ***	0.055	2.151 **	2.359 **
Delta_tilde_HAC	−10.909 ***	3.557 ***	2.685 ***	3.495 ***
Adjusted_Delta_tilde_HAC	−14.084 ***	4.592 ***	3.466 ***	4.511 ***

*, **, *** Indicates 10%, 5%, and 1% level of significance, respectively. (Delta_tilde and adjusted delta tilde [60]; Delta_tilde HAC and adjusted delta tilde HAC [61]).

. The null hypothesis of slope homogeneity was rejected except for the high-income group’s delta tilde and adjusted delta results. Thus, we estimated that our model allows slope heterogeneity.

Finally, CCEMG and AMG estimators were applied for each group separately, and detailed results are given in

Table 7. Results of the CCEMG and AMG. Source: Own Research.

	All Countries		High-Income
	CCEMG	AMG	CCEMG	AMG
HDI	−0.0194062	0.1401707 ***	−0.093534	0.0915567 *
	[0.1081822]	[0.0389843]	[0.092892]	[0.0524386]
	0.858	0.000	0.314	0.081
GII	0.0814868 ***	0.0876297 ***	0.0569406	0.0704541 *
	[0.0273326]	[0.0289037]	[0.040124]	[0.0364321]
	0.003	0.002	0.156	0.053
Constant	−0.0115027	0.0823168 ***	−0.044498	0.0767655 ***
	[0.041952]	[0.0229904]	[0.041175]	[0.0286834]
	0.784	0.000	0.280	0.007
RMSE	0.0823	0.1102	0.0649	0.0886
Wald	8.89 **	22.08 ***	3.25	8.73 **
p-value	0.0117 **	0.0000 ***	0.1973	0.0127 **
Observations	1210	1210	510	510
Number of countries	121	121	51	51
	Upper-Middle-Income		Lower-Middle-Income
	CCEMG	AMG	CCEMG	AMG
HDI	0.307407 **	0.1519242 **	−0.4413051	0.1673859 *
	[0.1505647]	[0.0592306]	[0.356378]	[0.100401]
	0.041	0.010	0.216	0.095
GII	0.0183536	0.0346355	0.1832053 **	0.179054 **
	0.0531783	0.0508424	[0.085822]	[0.0691064]
	[0.0531783]	0.496	0.033	0.010
Constant	0.0441121	0.1333106 ***	−0.0389114	0.0257806
	[0.063173]	0.0343834	[0.0959522]	[0.061434]
	0.485	0.000	0.685	0.675
RMSE	0.0831	0.1079	0.1115	0.1428
Wald	5.42 *	9.23 **	6.09 **	7.29 **
p-value	0.0667 *	0.0099 **	0.0477 **	0.0261 **
Observations	390	390	310	310
Number of countries	39	39	31	31

*, **, *** Indicates 10%, 5%, and 1% level of significance, respectively. Brackets denote z values.

. The results showed that only in high-income countries the model is not supported by the CCEMG estimator, while the rest of the analysis is supported. To clarify the interpretation of each group, evaluations should be considered separately. When all countries are assessed altogether, GII has a significant effect on GCI while HDI plays a significant role in GCI except for the CCEMG estimator. Moreover, for the high-income group, the coefficient of HDI for AMG is positively significant whereas GII is only positively significant for the AMG estimator. Furthermore, for upper-middle-income countries, results differentiate from other income groups since there is a positive relationship between HDI and GCI but there is not a positively significant relationship between GII and GCI. For lower-middle-income countries, results indicate that for the AMG estimator, there is a positive and significant effect of HDI and GII on GCI while HDI negatively affects GCI in the CCEMG estimator.

After examining the long-term relationship among the variables, we use Dumitrescu and Hurlin’s [55] panel causality test to determine the causal relationship between each pair of variables.

Table 8. Results of the Pairwise Dumitrescu Hurlin Panel Causality Tests.

	All Countries			High-Income
Null Hypothesis	W-Stat.	Zbar-Stat.	Prob.	W-Stat.	Zbar-Stat.	Prob.
GII → GCI	1.561 ***	4.360	0.000	1.284	1.435	0.151
GCI → GII	−2.1 × 1013 ***	−6.90 × 1013	0.000	1.438 **	2.212	0.027
HDI → GCI	2.596 ***	12.411	0.000	2.723 ***	8.702	0.000
GCI → HDI	7.308 ***	49.067	0.000	8.980 ***	40.298	0.000
HDI → GII	4.381 ***	26.301	0.000	5.787 ***	24.174	0.000
GII → HDI	1.528 ***	4.107	0.000	1.528 ***	2.665	0.008
	Upper Middle Income			Lower Middle Income
Null Hypothesis	W-Stat.	Zbar-Stat.	Prob.	W-Stat.	Zbar-Stat.	Prob.
GII → GCI	1.332	1.464	0.143	2.407 ***	5.541	0.000
GCI → GII	1.267	1.180	0.238	2.806 **	2.168	0.030
HDI → GCI	2.525 ***	6.734	0.000	2.484 ***	5.842	0.000
GCI → HDI	6.802 ***	25.621	0.000	4.970 ***	15.630	0.000
HDI → GII	3.795 ***	12.343	0.000	2.806 ***	7.111	0.000
GII → HDI	1.267	1.180	0.238	1.131	0.515	0.607

**, *** indicates 5%, 1% level of significance, respectively.

illustrates the causal relationships among variables. Specifically, there are bidirectional causal relationships between HDI ↔ GCI while there is a unidirectional relationship between HDI → GII for all income groups. Other causal relationships are related to the income groups. Some examples are as follows: There is not any causal relationship between GII and GCI in upper-middle-income countries. For high-income countries, there is no casual effect of GII on GCI. In lower middle-income countries, only GII affects GCI.

As a robustness test, we employ the Two-Way Fixed Effect test [65] to validate the outcomes of the CCEMG and AMG models, as presented in

Table 9. Robustness Test.

Variables	All Income	High Income	Upper Middle Income	Lower Middle Income
GII	0.0906 *** [4.82]	0.0931 *** [3.66]	0.0469 [1.36]	0.1568 [3.45]
HDI	0.16055 *** [4.03]	0.1999 ** [3.35]	0.1761 *** [3.38]	0.0605 [0.41]
Intercept	0.0910 *** [4.85]	0.0697 *** [2.75]	0.1185 *** [3.85]	0.0704 [1.51]
Country Dummies	No	No	No	No
Year Dummies	Yes	Yes	Yes	Yes
Observations	1210	510	390	310
F Stat	251.64	170.37	85.82	35.05
p-value	0.000	0.000	0.000	0.000

**, *** indicates 5%, 1% level of significance, respectively.

. The results affirm the robustness of the CCEMG and AMG findings. Notably, within the upper-middle-income group, the nonsignificant relationship between the GII and the GCI aligns consistently with the results. Similarly, the lack of significance between the HDI and GCI in lower-middle-income countries is sustained across the findings. Furthermore, the robustness test substantiates the significant relationships identified in other country groups.

5. Discussions

The results show that HDI has a positive significant effect on GCI in each income group which supports the findings of panel data studies proposed by Kiselakova et al. [51] and Hamid [41]. On the other hand, the impact of GII on GCI varies across income groups. In the upper-middle-income group, it is found that GII does not positively affect GCI which is supported by the Granger causality test [64]. In similar studies, regression analysis results show that GII has a positive impact on GCI [52,53]. Since we perform our analysis by dividing countries based on income groups, we only find a positive impact of GII on GCI in high-income and lower-middle-income countries. Therefore, this finding shows the importance of the diversion of income groups. The results of this study are consistent with prior research in the literature. In all-country analysis, human development has been found to have an impact on national competitiveness similar to Hamid’s [41] results. As nations aspire to bolster their global competitiveness, the facets encapsulating human development, namely education, life expectancy, and income, should be considered while explaining the significance of human development on competitiveness across all income groups of countries. Therefore, discerning suggestions of how each of these subcomponents might contribute to overall competitiveness is required. Designing an education system where it is compulsory up to a certain level and offering equal opportunity for all can be a first step. On the other hand, raising citizens’ living standards can be another good starting point to increase the competitiveness of a nation. As the quality of life boosts, life expectancy increases, too, and individuals try to accomplish the last step of Maslov’s hierarchy. Self-actualization also increases levels of creativity and productivity. Also, with the COVID-19 pandemic, we observed that despite being at the peak of globalization, each country must be capable of protecting its people. By providing quality health services that are accessible to all, nations can improve citizens’ life expectancy. These changes demonstrate itself as a cycle of growing influence from individuals to nations.

Furthermore, research by Cvetanovic et al. [53] in six Western Balkan countries (Albania, Bosnia and Herzegovina, Macedonia, Serbia, Croatia, and Montenegro) found no effect of innovation on competitiveness, which is consistent with the results we found for the same group as all countries except Serbia are in the upper-middle-income category. In the context of upper-middle-income countries, the translation of innovation into competitiveness might face several obstacles. Given the globalized nature of economies, these countries often encounter formidable competition from both high-income nations and lower-cost economies, which may challenge the establishment of a distinctive competitive edge. The challenge of effectively transferring technology from research to practical applications can also hinder the progression of innovative initiatives. Moreover, an underdeveloped entrepreneurship ecosystem, characterized by inadequate mentorship and funding avenues, may impede the transformation of novel ideas into marketable solutions. In some cases, a lack of robust local market demand or purchasing power could curtail the adoption of innovative products or services. Cumbersome regulatory processes and a risk-averse business environment may discourage the pursuit of innovative solutions, while insufficient collaboration between academic, industrial, and governmental entities could stifle the progress of innovative projects. Weak institutional support, encompassing factors like inadequate intellectual property protection and funding mechanisms, can obstruct the successful commercialization of innovation. To address these complex obstacles, a carefully tailored and comprehensive approach is necessary. This should include policy reforms, strategic investments in education and research, the development of collaborative networks, the establishment of supportive institutional frameworks, and the strategic alignment of innovation efforts with broader economic goals.

Additionally, due to their well-established research infrastructure, thorough educational systems, strong commitment to R&D investment, advanced technology transfer mechanisms, mature entrepreneurial ecosystems, and favorable regulatory environments, high-income countries frequently excel at converting innovation into competitiveness. These factors collectively foster a culture of innovation and its successful commercialization.

Even modest improvements in the innovation systems of lower-income countries can bring about noteworthy outcomes in the form of heightened competitiveness. Such incremental advancements have the potential to boost worker competence, stimulate better technology integration, streamline regulatory frameworks, and strengthen entrepreneurship-supporting ecosystems. These combined effects ultimately help these countries’ economies to grow more broadly and become more globally competitive. In conclusion, it is crucial to emphasize the enormous impact of innovation on the maintenance and improvement of competitiveness. While institutions must work to create a supportive infrastructure that fosters an environment favorable to innovation, policymakers are responsible for creating wise policies. Additionally, the effective conversion of innovative ideas into concrete results relevant to market dynamics and the commercial environment emerges as a key determinant.

6. Conclusions and Further Research

This study aims to investigate the effects of HDI and GII on GCI among various income groups between 2010 and 2019. A total of 121 countries are included in the analysis, and income groups are divided into three classes: 51 high-income countries, 39 upper-middle-income countries, and 31 lower-middle-income countries. There are two main goals of this research, first to investigate the relationship among these indices, and second to determine whether there is any difference in these relationships regarding income groups of nations. Except for the upper-middle-income group, it can be concluded that nations can increase their competitiveness by focusing on human development and creating an innovative environment. The contributions of this research are as follows. First, this is the first attempt to examine the effects of human development and innovation on competitiveness in the same model at the national level by using a panel data approach. Second, our results show that countries should be evaluated based on their income groups, not as a whole. Third, our results also show that upper-middle-income group countries behave differently than the other country groups regarding the effect of innovativeness on competitiveness.

Even though this study makes important contributions to the literature, it has some limitations that point to future lines of research. First, data availability is the main issue due to using secondary data derived from global indices. Additionally, the time span can also be updated and expanded. Moreover, due to a significant missing data problem, low-income countries are extracted from this study. Therefore, if data become available, the low-income group should be included. Finally, nation-related variables might be added to the model such as cultural characteristics or geographical locations to investigate country dynamics.