Measuring Inclusive Growth in Developing Countries: Composite Index Approach and Sectoral Transformation Analysis

Abstract

Inclusive growth is increasingly recognized as being critical to sustainable development, particularly in the context of rising income inequality and social polarization around the globe. Effective policy requires robust measurement, prompting the need to move beyond GDP and supplement traditional economic indicators. This study proposes a novel inclusive growth index (IGI) for 73 developing countries. The index is constructed using factor analysis with principal component analysis (PCA) across four pillars: economy, living conditions, equality, and governance. Our results reveal significant heterogeneity among developing countries, largely driven by variations in economic development and governance. Further analysis using OLS regression explores the impact of sectoral transformation, demonstrating a statistically significant positive relationship between shifts from the agricultural to the service sector and the IGI. These findings provide valuable insights for policymakers seeking to create more opportunities and target interventions to achieve more inclusive growth in developing economies.

1. Introduction

In recent decades, developing economies have striven to achieve inclusive economic growth through implementing reforms to alleviate poverty and decrease inequality. The dramatic reduction in poverty is highly associated with rapid economic growth and driven by high growth rates in a few countries such as China, India, and Viet Nam. Undoubtedly, some level of economic growth is necessary for poverty reduction and economic progress, but growth by itself is not sufficient for the sustainable and equitable improvements of welfare (Ali & Son, 2007; Sergi et al., 2023). Economic growth does not guarantee that people benefit equally (Ahsan et al., 2024; Popov, 2025) and that dividends are shared fairly among social groups (Boarini et al., 2015). Economic growth alone does not ensure sustained poverty reduction and inequality decrease, which threatens social cohesion and can undermine growth.

Many studies affirm that economic growth has resulted in growing inequality (Heshmati et al., 2019; Popescu, 2025), which, in turn, lowers the influence of growth on poverty reduction and growth itself. In this context, rising inequality has become a main concern for policymakers; therefore, more attention has been drawn to inclusive growth. Although the dominant development discourse focuses on the "growth-first" paradigm, government policies emphasize reducing inequality by creating pro-poor growth, providing social protection mechanisms, delivering public services, and establishing institutional conditions that foster inclusive and equitable development and financial inclusion. (Cook, 2006; Seti et al., 2025; Mishra et al., 2024).

The grand theory of inclusive growth is based on integrating economic growth with equity, participation, and sustainability. It emphasizes that growth must be broad-based and include all segments of society. Additionally, it must be linked to reducing poverty and expanding social opportunities. This theory combines key concepts from human capital theory, social capital theory, development economics, and institutional economics. Currently, this theory is reflected in the approaches of international organizations such as the World Bank, the Asian Development Bank, the OECD, the IMF, the UNDP, and the G20, which emphasize balanced sectoral development, productive employment, poverty reduction, and social inclusion as integral to growth. Inclusive growth has also gained prominence in the discourse on UN development policy. The 2030 Agenda, the Sustainable Development Goals (SDGs), and national strategies focus on paving the path to sustainable development and establishing solid foundations for inclusive growth.

To achieve this development objective, it is necessary to systematically measure inclusiveness. A quantitative assessment of policy outcomes is crucial for clarifying targets, refining guidance for achieving inclusive growth, and identifying deviations from objectives.

The evaluation of inclusive growth has always focused on the scope of interest for both researchers and international organizations. Inclusive growth is a broad and multidimensional phenomenon, which makes it difficult to have a universal measure of inclusiveness. In order to assess countries’ performance in inclusive growth, it is becoming vital to construct an integral index of inclusive growth that is not only a working tool for updating benchmarks but that also provides guidance, reframes objectives, reorders priorities, and makes social policy adjustments. Currently, various sets of indicators and methodological approaches are being developed by different international organizations such as UNCDAT, the World Economic Forum, etc. They are all based on the same logic of building a composite index of inclusiveness through a set of indicators describing various aspects of inclusive growth and integrating them into one composite index. All these approaches are within the framework of economic logic; however, from a methodological perspective, they are limited and problematic. Finding solutions to those issues is necessary in order to objectively reflect the outcomes of policies on inclusive growth, unveil opportunities and impediments, and develop and implement strategic approaches toward inclusive growth.

Over the past decade, developing economies have shifted away from the non-productive agricultural sector to productive non-agricultural sectors, including reallocations of the workforce across sectors. This kind of economic transformation can increase productivity and reduce poverty and income inequality. This type of structural transformation may be called inclusive transformation. While the existing literature explores the relationship between structural transformation and poverty and income inequality (Sergi et al., 2023; Seti et al., 2025; Mishra et al., 2024; Popescu, 2025), these studies often focus on specific aspects of inequality or utilize limited measures of development, leaving a gap in our understanding of the overall impact on inclusive growth. Therefore, there is a lack of research on the impact of economy transformation regarding a more holistic inclusive growth index.

The purpose of this paper is twofold: First, we develop a composite index of inclusive growth based on principal component analysis (PCA) using a wide range of indicators that capture different dimensions of inclusiveness in developing countries. Second, we aim to assess the impact of structural change on this inclusive growth index by applying ordinary least squares (OLS) regression models.

Therefore, this study contributes to the literature by providing a more comprehensive measure of inclusive growth and examining the direct impact of structural transformation on this index, offering insights into the policies that can maximize the inclusive growth benefits of economic development.

This paper is organized into six sections. Section 1 introduces this article’s research area. Section 2 presents a literature review and briefly discusses different approaches to inclusive growth assessment. Section 3 gives an overview of the methodology applied in this study. The results are presented in Section 4. In Section 5, the methodological issues and their solutions are revealed. Section 6 is dedicated to the concluding reviews.

2. Literature Review

Inclusive growth is a development agenda that deals with poverty reduction, the mitigation of inequality, and sustainable development. There is no agreed-upon and common definition of inclusive growth; however, it refers to growth accompanied by equal opportunities and contains economic, social, and institutional aspects concentrating on making opportunities accessible to all. Inclusive growth supposes that all members contribute to growth equally and that the economic opportunities created by growth are available to all, especially the poor.

Inclusive growth is a multidimensional and complex phenomenon that refers to the enhancement of everyone’s well-being and quality of life equally (Barnat et al., 2023). Inclusive growth contributes to poverty reduction and allows people to benefit from economic growth. It is in line with the absolute definition of pro-poor growth, supposing that poor people benefit from growth. Therefore, rapid growth is essential for poverty reduction, but long-term, sustainable growth is required to be broad-based across sectors and inclusive of most of the national labor force. This definition emphasizes the significance of structural transformation for economic diversification and competition. As a main tool for inclusive growth, productive employment is discussed (Ianchovichina & Lundstrom, 2009).

Mohazzam and Atif highlighted that economic growth resulted in better education, improved health services, increased availability of health services, and increased employment opportunities, generating effective human capital and high-skilled labor. The authors affirmed that for inclusive and pro-poor growth, the focus of public expenditure should be revisited by investing more in health and education (Atif & Mohazzam, 2012).

Inclusive growth is broadly perceived as broad-based equitable growth, economic growth with human development, pro-poor growth, accessible and participatory growth, and financially and environmentally sustainable growth (Mitra & Das, 2018). The concept of inclusiveness is based on the provision of productive employment to enhance economic opportunity and equal access to economic opportunities through investment in human capital and social mechanisms to protect poor and vulnerable social groups (Hosono, 2022).

The World Bank, in an earlier approach, considered inclusive growth as a rapid pace of economic growth that is essential to reduce absolute poverty and that involves a large share of the labor force to be sustainable, emphasizing productive employment. Recently, the notion of shared prosperity has been more widely used to refer to income growth among the bottom 40 percent of the population (World Bank, 2022). The OECD defines inclusive growth as economic growth that provides opportunities for all social groups and that shares the benefits of increased prosperity equally across society. Moreover, this approach to economic growth goes beyond income and includes non-monetary dimensions of well-being such as employment prospects, health outcomes, and educational opportunities (OECD, 2015). According to UNDP, inclusive growth refers to equity with growth or shared prosperity from economic growth. Inclusive growth supposes that everyone can participate in the growth process and that its benefits are shared equitably (United Nations Development Programme, 2017). While defining inclusive growth, the Commission on Growth and Development emphasized the equality of opportunity in terms of access to markets, health, education, and other resources, giving everyone the option to enjoy the fruits of growth (Commission on Growth and Development, 2008).

The Asian Development Bank (ADB) adopted inclusive growth as an essential development agenda in parallel with sustainable growth. Among the recommendations by the Asian Development Bank literature, the authors outlined the following essential measures that should be taken to achieve inclusive growth: fostering efficient and sustainable economic growth, ensuring a level political playing field, enhancing capacities, and establishing safety nets (Rauniyar & Kanbur, 2010).

The measurement of inclusive growth has recently been the focus of economists’ and international organizations’ research. There is a wide range of indices with different approaches. It should be noted that the multifaceted nature of inclusive growth makes it hard to try to measure its main characteristics. There is no common set of indicators based upon which to measure the inclusive growth of countries. So far, most indicators of economic performance have been used; however, restrictions on GDP reveal the need to complement GDP with other non-monetary indicators. As a result, international organizations have developed various composite indices, which are useful in understanding the impact of policies and reviewing mechanisms.

The methodology of the World Economic Bank and the World Economic Forum was applied to measure the index of inclusive growth for 26 regions of the Russian Federation and throughout the country as a whole. The authors used key indicators of inclusive growth and development performance divided into three groups: growth and development, inclusion, and intergenerational equity and sustainability. In addition to the World Economic Forum methodology, the authors also included an indicator of purchasing power. The results were used to define the strengths and weaknesses of inclusive growth and development that could be later used as a basis for the formation of inclusive growth and development policies (Sharafutdinov et al., 2019).

The European Commission outlined three priorities within the Europe 2020 Strategy: smart, sustainable, and inclusive growth. Inclusive growth was defined as a high-employment economy that delivers economic, social, and territorial cohesion, creating equal opportunities for people through high employment levels, investment in skills development, poverty reduction, and improved labor markets and social protection systems. Five targets were set for all three priorities based on the following key indicators: employment rate; R&D investment (% of GDP); climate and energy targets, including greenhouse gas emissions, renewable energy shares, and energy efficiency improvement; the share of early school leavers and tertiary education attainment; and people at risk of poverty (European Commission, 2010).

The World Economic Forum calculated an inclusive development index based on 12 indicators that assessed countries’ progress in inclusive growth. These indicators were grouped into three pillars: growth and development; inclusion; and intergenerational equity and sustainability. The first pillar was represented by indicators such as GDP per capita, employment rate, labor productivity, and healthy life expectancy. The second pillar was measured using median household income, poverty rate, and income and wealth Gini indicators. The metrics for intergenerational equity and sustainability included adjusted net savings, public debt as a percentage of GDP, the dependency ratio, and the carbon intensity of GDP (World Economic Forum, 2018).

The Asian Development Bank’s publication “Framework of Inclusive Growth Indicators” defined inclusive growth as economic growth with equality of opportunity and proposed 35 inclusive growth indicators for its 48 regional member economies (Asian Development Bank, 2014). The framework of inclusive growth indicators included the following:

• Poverty and inequality indicators (including monetary and non-monetary dimensions): indicators of income, years of schooling, and child health.
• Growth and the expansion of economic opportunities: economic growth, employment, and infrastructure indicators.
• Social inclusion to ensure equal access to economic opportunities, including indicators on access to education and health, basic infrastructure, utilities and services, and gender equality and opportunity.
• Social safety nets: social protection and social security expenditure indicators.
• Good governance and institutions: indicators of voice and accountability, government effectiveness, and the control of corruption.

The OECD outlined 24 indicators grouped into four categories to measure the progress of countries in inclusive growth (OECD, 2018):

• Growth and ensuring the equitable sharing of benefits: These indicators were used to monitor whether the economy was growing and whether standards of living were improving for different social groups.
• Inclusive and well-functioning markets: These indicators provided insights into the structure and functioning of the economy and marketplaces, monitored the effectiveness of product and labor markets, and revealed the relationship between productivity and inclusiveness.
• Equal opportunities and foundations of future prosperity: These indicators were focused on non-monetary components of well-being, describing people’s opportunities regarding health, education, environmental quality of life, and childcare.
• Governance: These indicators revealed the effectiveness and responsiveness of the government.

Furthermore, the United Nations Conference on Trade and Development (UNCTAD), in collaboration with the Eurasian Economic Commission (EEC), developed a composite index to measure inclusive growth in the region. The authors highlighted the methodological issues they faced when developing such a composite index, which were common and not specific to EEC countries. In an attempt to move beyond GDP, the authors used different concepts and measures, and 21 indicators were ultimately selected as the most relevant for inclusive growth, grouped into three pillars: economics, living conditions, and inequality. Principal component analysis was then used to assign weights. The authors outlined the importance of these results for policymakers in paving the way for inclusive growth (Barnat et al., 2023).

Previously, Ali and Son (2007) proposed another approach to measuring inclusive growth based on increasing the social opportunity function, which considered two factors: the average opportunities available to the population and the distribution of opportunities among the population. The authors studied the case of the Philippines, applying the opportunity curve to analyze the inclusiveness of growth. By analyzing the dynamics of the opportunity curve over time and the specific income distribution segments where shifts occurred, judgments could be made about the degree of inclusiveness.

Cichowicz and Rollnik-Sadowska measured the level of inclusive growth among Central and Eastern European countries using inclusive indicators grouped into seven pillars, applying principal component analysis. A comparative analysis of the years 2006 and 2016 revealed differences in countries’ performance in inclusive growth and the structure of factors. For instance, in 2016, the inclusive growth variables were categorized under the factors of “Education expenditures, unsatisfactory medical services, and state transparency”, “Social inclusion”, and “Labor market situation”, whereas, in 2006, they were categorized under “State-created conditions for quality of life”, “Social inclusion”, and “Labor market situation and digitalization”. The authors attributed these differences to the specific factors of each country, the advantages obtained from EU membership, and the countries’ resilience to the 2007 global economic crisis (Cichowicz & Rollnik-Sadowska, 2018).

Mitra and Das proposed an inclusive growth index for 16 Asian countries based on 24 developmental indicators, as an integral measure of inclusive growth. The authors used an ad hoc weighting scheme and a weighting scheme based on principal component analysis (PCA) to construct the index (Mitra & Das, 2018).

Lin developed the inclusive sustainable transformation index, which measured the country’s progress made toward an inclusive and environmentally friendly modern economy. The index was calculated for almost 200 countries over 25 years and helped policymakers identify policy areas that succeeded and gaps that required more efforts (Lin et al., 2019).

Ahsan and co-authors affirmed that inclusive growth is both the outcome and the process as it emphasizes that poor people both participate in the economic growth process and benefit from growth equally. Therefore, it was concluded that policymakers should direct their efforts to enhancing human capacities rather than pro-poor growth policies. A composite index of inclusive growth was constructed via principal component analysis utilizing 24 different socio-economic variables grouped into seven dimensions: economic growth, productive employment, infrastructure endowment, human capabilities, income inequality, gender equity, and governance. Moreover, the inclusive growth index calculated from 1980 to 2022 was used in a linear regression model as a dependent variable with foreign direct investment, financial deepening, inflation rate, and domestic investment as explanatory variables in the case of Pakistan. The results showed that inflation and domestic and foreign investments were significant determinants of inclusive growth in Pakistan (Ahsan et al., 2024). The results were similar to those of a study previously conducted by the IMF that affirmed that macroeconomic stability, human capital, and structural changes are the foundations for inclusive growth. In terms of structural change, trade openness and foreign direct investment contribute to inclusive growth, whereas financial deepening has no discernible impact (Anand et al., 2013).

Saad developed a robust measure of inclusive growth based on key determinants such as education, health, income distribution, and environmental sustainability, applying principal component analysis. Afterward, a generalized method of moments (GMM) was used to observe the impact of public spending on fostering inclusive growth in developing countries. The results showed that targeted investments in education, health, and the public good positively affected inclusive growth, whereas high unemployment rates had a negative influence (Saad, 2024).

Ghosh developed a composite inclusive growth index (IGI) based on fifteen indicators on economic expansion, environmental sustainability, gender equity, human capabilities, and financial inclusion for India over the period 1990–2020. A two-stage principal component analysis (PCA) was applied to define the weights of the indicators. The inclusive growth index was useful to evaluate the country’s performance and achievements in inclusive growth and to discover improvement rates for all dimensions (Ghosh, 2023).

Structural transformation, driven by innovation and digitalization, has also been observed as an essential factor with an impact on inclusive growth. However, behind the positive influence of structural transformation in fostering economic growth, it can deteriorate income inequality. Mamman and Sohag applied quantiles via a moment panel model to examine the impact of structural transformation for all African countries. The results affirmed that the structural transformation from agriculture to services reduced poverty and increased inequality; however, digitalization and technological processes reduced both extreme poverty and inequality (Mamman & Sohag, 2023).

Armah and co-authors studied the relationship between structural changes and the economic, social, and political aspects of inequality. The authors outlined that structural transformation is not a sufficient condition for inclusive growth and that, without appropriate policies, it can lead to rising inequalities (Armah et al., 2014).

Policymakers initiate structural reforms to provide inclusive growth, and the growth-enhancing combination and sequences of reforms are the most important issues. One study (Ari et al., 2022) analyzed the impact of structural reforms implemented in 126 countries between 2000 and 2019 on economic growth and confirmed that structural reforms in the area of product, labor, and financial markets as well as the legal system had a significant impact on economic growth in a 5-year period (Ari et al., 2022).

Alvarez-Cuadrado and Poschke introduced the “push” and “pull” hypothesis to describe drivers of structural change in the economy. According to the authors, advancements in agricultural technology shift resources to industries, and advancements in the industrial sector boost the wages of industry workers, pushing the labor force into that sector (Alvarez-Cuadrado & Poschke, 2011).

According to Swiecki, in developed countries, the driving force of the decline in the share of the labor force in the manufacturing sector and the growth in services is sector-biased technological advancement. In poorer countries, the shift of the labor force away from the agricultural sector is due to non-homothetic preferences (Swiecki, 2017).

The correlation between the sectoral structure of the economy and income distribution has been well analyzed by economists. One study (Mesa Salamanca & Zuleta Gonzalez, 2021) investigated the inverse correlation and found a negative relationship between income inequality and the share of the agricultural sector in middle-income and developing countries. The relationship was positive in developed countries.

Transformations of the economy may deepen the gap between skilled–unskilled wage earners. One study (Pi & Zhang, 2018) found that the impact of the change in the urban skilled sector on wage inequality depended on the capital–labor ratio in the sector.

Investigating the inclusiveness rate of the economy of Thailand, Warr and Suphannachart analyzed medium-to-long-term correlations between structural transformations, poverty rates, and income inequality in Thailand. The authors concluded that the state of the economy depended on the chosen definition of the term “inclusive growth”. They differentiated between inclusive growth, reducing inequality, and reducing poverty. According to the authors’ investigation during the period of 1981–2017, Thailand’s economy was inclusive since, during the whole period, inequality and poverty declined. The poverty-reducing power of growth depended on the poverty line chosen (Warr & Suphannachart, 2022).

Employing dynamic equilibrium model, Wang analyzed the interaction between industrialization and rural income distribution. Increases in the production of high-demand goods can boost economic growth, but this can also be slowed down by decreases in demand due to income inequality (Wang, 2019).

Mihaylova and Bratoeva-Manoleva proved that in the Bulgarian economy, the relationship between changes in the structure of the economy and wage inequality was bilateral. They proved that the shift of the economy from agricultural and industrial sectors to the service sector was due to the wage gap between the sectors of the economy, regions of the country, and levels of education, which, in turn, triggered wage inequality (Mihaylova & Bratoeva-Manoleva, 2018).

Investigating the poverty-reducing power of the structural change in the economy Erumban and de Vries used the GGDC/UNU-WIDER Economic Transformation Database for 42 developing countries for the period of 1990–2018, employing regression analysis. They showed that the increase in the productivity of the manufacturing sector triggered a decrease in the poverty rate in developing Asian countries (Erumban & de Vries, 2024). The same data were used by Kruse and co-authors to analyze employment trends in the manufacturing sector in developing economies, employing the method of regression analysis (Kruse et al., 2023).

The poverty-reducing power of economic transformation was well analyzed by Rifa and Listiano (Rifa’i & Listiono, 2021) using panel data for 38 cities of Indonesia employing the OLS modeling technique. To analyze the relationship, Enongene (Enongene, 2024) applied value-added sectoral analysis for Sub-Saharan African countries. Ardiansyah and co-authors employed the VECM model with sectoral added-value data in Indonesia to examine the nexus (Ardiansyah et al., 2020). Value-added proportion data with ARDL analysis was also implemented by Mateko (Mateko, 2025). The same logic was used by Frikha and Gabsi to examine the impact of structural transformation on poverty reduction in 13 emerging countries (Frikha & Gabsi, 2024).

Employing a regression model for the Groningen Growth and Development Center data of 42 developed and developing countries, Rodrik concluded that deindustrialization in developing economies would lead them to invent new, service-led growth models (Rodrik, 2016). The main sources of the transition were well-designed institutions and the growth of skilled human capital.

In summary, the literature review suggests that despite the various approaches adopted by researchers and international organizations to conceptualize inclusive growth, its measurement remains a subject of debate. Despite numerous studies, there is no universal consensus on the range of indicators or methodological approaches for evaluating inclusive growth. Furthermore, there are no all-encompassing determinants of inclusive growth that vary from country to country, nor any agreed-upon method of assessing the impact of structural economic changes on inclusive growth. While it is broadly recognized that structural economic transformation may positively affect poverty and income inequality, the relationship between structural economic change and the overall inclusiveness of the economy has not been well examined.

3. Methodology and Data

Based on the literature review, different aspects of inclusive growth grouped under four pillars were distinguished. For the most comprehensive assessment of inclusive growth according to data availability, the selection of indicators was important. Wide-ranging indicators for the inclusive growth index were chosen as they mostly covered all aspects of inclusive growth and have been intensively utilized by researchers and international organizations to develop indices of inclusiveness.

The first pillar, economy, included key indicators of economic output, productivity, exports, and poverty rates to reveal the level of economic development. The second pillar, living conditions, contained essential indicators such as life expectancy rates, school enrollment, damages caused to the natural environment by economic production, and adjusted net savings, excluding particulate emission damage. The third pillar, equality, included indicators reflecting income and wealth equalities, equality in labor participation, and gender equality. The fourth pillar, governance, evaluated the quality of governance and measured the countries’ progress on establishing public institutions, democracy, political stability, effective governance, rule of law, and the regulatory environment.

Hence, the above-mentioned 23 indicators under the above-mentioned four pillars were identified as being relevant to inclusive growth, considering both socio-economic issues, gaps in inequality, governance indicators, and environmental indicators. Those indicators were also matched or linked to indicators involved in the SDG indicator framework. The cross-sectional data of 73 countries in 2023 were compiled from multiple sources such as the World Inequality Database by the World Inequality Lab, Worldwide Governance Indicators, World Development Indicators Database of the World Bank, and the ILOSTAT Database of the International Labour Organization (ILO, n.d.) (

Table 1. The indicators of the pillars of the inclusive growth index (IGI) and their sources.

Pillar	Indicators	Abbreviation	Source
Economy	GDP per capita (constant 2015 USD) Output per worker (GDP constant 2021 international USD at PPP) Exports of goods and services (% of GDP) Poverty headcount ratio at USD 3.65 a day (2017 PPP) (% of population)	GDP_per_capita Output_per_worker Export_to_GDP Poverty_rate_USD 3.65	(World Bank, n.d.-a) (ILO, n.d.)
Living Conditions	Life expectancy at birth Secondary school enrollment (% gross) Adjusted net savings, excluding particulate emission damage (% of GNI) Carbon intensity of GDP (kg CO2e per 2021 PPP USD of GDP)	Life_expectancy_at_birth School_enrollment Adjusted_net_savings_2021 Carbon_intensity_2021_PPP	(World Bank, n.d.-a) (ILO, n.d.)
Equality	Post-tax national income Gini coefficient Post-tax national income top 1% Post-tax national income top 10% Net personal wealth Gini coefficient Employment rate (females), % Employment rate (males), % Youth unemployment rate (aged 15–24) NEET rate as the percentage of the population of young people aged 15–24 neither in employment nor in education or training Proportion of seats held by women in national parliaments (%) Women business and the law index score	Post_tax_Gini_coefficient Income_top_1 Income_top_10 Wealth_Gini Employment_female Employment_male Unemployment_youth NEET Seats_women__in_parliament Women_business_and_law	(World Inequality Lab, n.d.)
Governance	Rule of law Government effectiveness Voice and accountability Regulatory quality Control of corruption	Rule_of_law Gov_effectiveness Voice_and_account Regulatory_quality Control_of_corruption	(World Bank, n.d.-b)

Source: developed by the authors based on the World Bank (n.d.-a); ILO (n.d.); World Inequality Lab (n.d.); and World Bank (n.d.-b).

). Countries were chosen according International Monetary Fund (IMF) classification.

The first stage of the construction of the inclusive growth index (IGI) was the normalization of 23 variables using Z-scores. Normalization converted the values with different scales to a common scale with a mean of zero and a standard deviation of one. The Z-score was calculated as

Z_ij = (X_ij − M_j)/S_j

(1)

where X_ij is the value of the ith indicator in the jth country, M_j is the mean value of the ith indicator, and S_j is the standard deviation of the ith indicator.

It should be noted that for the poverty rate, Gini coefficients, national income top 1% and 10%, NEET, unemployment rates, and carbon intensity of GDP indicators, Z_ij was considered.

After normalization, principal component analysis (PCA) was conducted using SPSS Statistics 21 software (Version 21.0.0.0.) to construct an inclusive growth index. PCA-based factor analysis reduces a large number of variables to a smaller set of factors that are meaningful and interpretable. The objective was to analyze the dependence structure, which made the characteristics of that structure more simplified and clear. In other words, as an extraction method, PCA allowed us to reduce the dimensionality, reveal latent structures, and identify underlying components among the observed variables.

Before performing the analysis, the Kaiser–Meyer–Olkin (KMO) test was conducted as a measure of sampling adequacy while assessing the goodness of fit. The KMO values were greater than 0.5, which confirmed that the dataset was adequate for factor analysis. Bartlett’s test of sphericity was also carried out; the null hypothesis of this test was that the variables were not correlated. The result of the p-value was 0.000, which confirmed that the variables were significantly correlated. The results of both tests indicated that there was some degree of correlation among the variables and that the data was adequate for principal component analysis.

At the next stage, the extraction of principal components within each pillar was conducted based on eigenvalues greater than one criterion. According to this methodological approach, the first factor explained the largest amount of total variance in the data, the second factor was extracted to determine the greatest share of the remaining variance, and the process continued accordingly. It should be noted that for the complete description of the observed variables, the factor extraction procedure continued as long as the eigenvalues of each subsequent principal component remained greater than 1.

Once factors were extracted, Varimax rotation was used as an orthogonal rotation method to perform the factor loadings that represented the correlations between the variables and factors. The rotated component matrix showed a factor loading for each variable that revealed the importance of each variable to the component. It contained significant information regarding the structure of each factor that could be used for defining the weights of the observed variables. So, based on factor loadings, the weight of each variable was obtained within each pillar. Hence, the score of each pillar was constructed by multiplying Z-scores of indicators by the generated components’ loadings, considering the correlation of indicators within each component. The number of principal components retained in each pillar varied. If the pillar had a few factors, the ratio of the variance of each component to the total variance was considered. Afterward, the pillar score was standardized using the min–max formula:

$$Pilla{r}_{score}={\displaystyle \frac{X-X_{\min }}{X_{\max }-X_{\min }}}$$

(2)

where $X_{\max }$ and $X_{\min }$ are the maximum and minimum values of the non-standardized pillar score.

At the final stage, the inclusive growth index was constructed based on pillar scores, with equal weighting given to each pillar. This approach was chosen to avoid huge score volatility while applying PCA to the pillars and subjective expert bias. This implied that each pillar was equally significant in constructing inclusive growth.

To analyze the potential impact of the structural transformation of developing economies on inclusiveness, the OLS regression technique was employed. The model utilized data from the same 73 developing countries in 2023 which the inclusive growth index was calculated for. Due to missing data for some of the countries, the number of observations in the model declined to 64. The data used in this study came from the World Bank Databank/World Development Indicators. The inclusive growth index, developed by the authors, was used as a measure of inclusive growth. For the cross-sectional analysis, the logged value of the inclusive growth index was taken as the dependent variable.

The authors modeled the potential influence of both sectoral added value and sectoral employment shifts on the independent variable. In the first and second parts of the model, the respective logged values of sectoral percentage shares of GDP and sectoral employment shares were used as explanatory variables.

Firstly, we analyzed how the change in the shares of the three sectors affected the inclusive growth index. Secondly, the influence of the change in sectoral employment on the inclusive growth index was analyzed.

For the first part of the analysis, the following econometric models were developed:

Log_IGI = α + β1 × log_VAA + β2 × log_VAI + β3 × log_CPI + β4 × log_GFCF + ε

(3)

Log_IGI = α + β1 × log_VAA + β2 × log_VAS + β3 × log_CPI + β4 × log_GFCF + ε

(4)

Log_IGI = α + β1 × log_VAI + β2 × log_VAS + β3 × log_CPI + β4 × log_GFCF + ε

(5)

where IGI is the inclusive growth index, VAA is the value added in the agricultural sector (% of GDP), VAI is the value added in the industrial sector (% of GDP), VAS is the value added in the service sector (% of GDP), CPI is the consumer price index, GFCF is the gross fixed capital formation (% of GDP), and ε is the random error term.

Equations (3), (4) and (5), respectively show the potential impact of agriculture–service and industry–service shifts, agriculture–industry and service–industry shifts, and industry–agriculture and service–agriculture shifts on inclusive growth. The logarithmic values of all variables were taken.

For the second part of the analysis, the following econometric models were developed:

Log_IGI = α + β1 × log_EA + β2 × log_EI + β3 × log_CPI + β4 × log_GFCF + ε

(6)

Log_IGI = α + β1 × log_EA + β2 × log_ES + β3 × log_CPI + β4 × log_GFCF + ε

(7)

Log_IGI = α + β1 × log_EI + β2 × log_ES + β3 × log_CPI + β4 × log_GFCF + ε

(8)

where IGI is the inclusive growth index, EA is employment in agriculture (% of total employment), EI is employment in the industrial sector (% of total employment), ES is employment in the service sector (% of total employment), CPI is the consumer price index, GFCF is the gross fixed capital formation (% of GDP), and ε is the random error term. The logarithmic values of all variables were taken.

Equations (6)–(8), respectively show the potential impact of agriculture–service and industry–service employment shifts, agriculture–industry and service–industry employment shifts, and industry–agriculture and service–industry employment shifts on inclusive growth.

4. Results

For the first pillar, economy, only one principal component was extracted. This principal component was interpreted as the factor “economic development” and accounted for 71.08% of the total variance of the four original indicators involved in the pillar (

Table 2. Retained factor economic development for the pillar economy.

Component	Initial Eigenvalues			Extraction Sums of Squared Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	2.843	71.080	71.080	2.843	71.080	71.080
2	0.683	17.070	88.150
3	0.385	9.615	97.765
4	0.089	2.235	100.0
Component Matrix for First Component Extracted
Zscore (GDP_per_capita)			0.909
Zscore (Output_per_worker)			0.946
Zscore (Export_to_GDP)			0.647
ZPovertyRev			0.839

Source: developed by the authors based on the World Bank (n.d.-a); ILO (n.d.); World Inequality Lab (n.d.); and World Bank (n.d.-b).

).

For the second pillar, living conditions, two principal components were obtained that captured 74.622% of the total variance of the four original indicators included in this pillar. Life expectancy at birth and secondary school enrollment had strong correlations with the first principal component, which could be interpreted as the factor “human capital”. The second component had the strongest relationship with adjusted net savings and carbon intensity and could therefore be perceived as the factor “sustainability”.

Table 3. Retained principal components for the pillar living conditions.

Component	Initial Eigenvalues			Extraction Sums of Squared Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	1.858	46.442	46.442	1.857	46.428	46.428
2	1.127	28.180	74.622	1.128	28.194	74.622
3	0.785	19.613	94.235
4	0.231	5.765	100.0
Rotated Component Matrix *
Variables			1 component		2 component
Zscore (Life_expectancy_at_birth)			0.918		0.112
Zscore (School_enrollment)			0.908		−0.120
Zscore (Adjusted_net_savings_2021)			0.284		0.776
ZCarbonintensityRev			−0.331		0.706

* Rotation Method: Varimax with Kaiser normalization. Source: developed by the authors based on the World Bank (n.d.-a); ILO (n.d.); World Inequality Lab (n.d.); and World Bank (n.d.-b).

shows that “human capital” accounted for 46.428% of the total variance and “sustainability” explained 28.194 per cent of the total variance (Table 3).

For the third pillar, equality, three principal components were retained that accounted for 75.544% of the total variance of the observed ten indicators that were chosen to measure equality. The first component was highly correlated with normalized values of inequality indicators (post-tax national income and net personal wealth Gini coefficients, post-tax national income top 1%, post-tax national income top 10%); therefore, it could be named as the factor “income and wealth equality”, a measure of the absence of income and wealth disparities among different social groups. The second component measured the imbalances and gender gaps in the labor market and revealed the youth participation in work and education; therefore, it could be interpreted as the factor “labor participation equality”. The third component had a strong correlation with the proportion of seats held by women in national parliaments, women business, and the law index score indicators of pillar 3 and reflected gender involvement; therefore, it could be perceived as the factor “gender equality”. The factor “income and wealth equality” contributed to 31.903% of the variability for the four original variables of inequality. The second (“labor participation equality”) and third (“gender equality”) components explained 25.841 per cent and 17.799% of total variance, respectively (

Table 4. Retained principal components for the pillar equality.

Component	Initial Eigenvalues			Extraction and Rotation Sums of Squared Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	3.197	31.967	31.967	3.197	31.967	31.967
2	2.643	26.426	58.393	2.643	26.426	58.393
3	1.715	17.151	75.544	1.715	17.151	75.544
4	0.961	9.614	85.158
5	0.496	4.959	90.117
6	0.406	4.057	94.174
7	0.243	2.426	96.600
8	0.189	1.887	98.488
9	0.133	1.330	99.818
10	0.018	0.182	100.000
Rotated Component Matrix *
Variables			1 component	2 component	3 component
ZPost_tax_Gini_coefficientRev			0.946	−0.101	0.040
ZIncome_top_1Rev			0.862	−0.041	0.064
ZIncome_top_10Rev			0.985	−0.004	0.020
ZWealth_GiniRev			0.590	0.286	−0.448
ZNEETRev			0.334	0.564	0.290
ZUnemployment_youthRev			0.022	0.878	−0.134
Zscore (Employment_female)			0.003	0.794	0.431
Zscore (Employment_male)			−0.283	0.865	−0.151
Zscore (Seats_women__in_parliament)			−0.098	−0.028	0.729

* Rotation Method: Varimax with Kaiser normalization. Source: calculated by the authors based on the World Bank (n.d.-a); ILO (n.d.); World Inequality Lab (n.d.); and World Bank (n.d.-b).

).

Finally, for the fourth pillar, governance, only one principal component (“effective governance”) was extracted, which captured 82.442 per cent of the total variance of the five variabilities included in the pillar (

Table 5. Retained factor economic development for the pillar governance.

Component	Initial Eigenvalues			Extraction and Rotation Sums of Squared Loadings
	Total	% of Variance	Cumulative %	Total	% of Variance	Cumulative %
1	4.122	82.442	82.442	4.122	82.442	82.442
2	0.470	9.402	91.844
3	0.223	4.455	96.300
4	0.104	2.086	98.386
5	0.081	1.614	100.000
Component Matrix
Variables			1 component
Zscore (Gov_effectiveness)			0.908
Zscore (Regulatory_quality)			0.934
Zscore (Voice_and_account)			0.809
Zscore (Rule_of_law)			0.952
Zscore (Control_of_corruption)			0.931
Zscore (Gov_effectiveness)			0.908
Zscore (Regulatory_quality)			0.934

Source: calculated by the authors based on the World Bank (n.d.-a); ILO (n.d.); World Inequality Lab (n.d.); and World Bank (n.d.-b).

).

According to the factor loadings identified through PCA-based factor analysis, the weights of the original indicators within each pillar were assigned (see

Table 6. The weights of the original indicators within each pillar retained through PCA.

Pillar	Indicator	Weight
Economy	Zscore (GDP_per_capita)	0.272074
	Zscore (Output_per_worker)	0.283149
	Zscore (Export_to_GDP)	0.193655
	Zscore (Poverty_rate_USD 3.65)	0.251122
Living Conditions	Human Capital	0.6221758
	Zscore (Life_expectancy_at_birth)	0.502738
	Zscore (School_enrollment)	0.497262
	Environmental Quality	0.377824
	Zscore (Adjusted_net_savings_2021)	0.523617
	Zscore (Carbon_intensity_2021_PPP)	0.476383
Equality	Income and wealth Equality	0.422316
	Zscore (Post_tax_Gini_coefficient)	0.27963346
	Zscore (Income_top_1)	0.25480343
	Zscore (Income_top_10)	0.29116169
	Zscore (Wealth_Gini)	0.17440142
	Labor participation Equality	0.34207
	Zscore (NEET)	0.181877
	Zscore (Unemployment_youth)	0.283134
	Zscore (Employment_female)	0.256046
	Zscore (Employment_male)	0.278942
	Gender Equality	0.235614
	Zscore (Seats_women__in_parliament)	0.460227
	Zscore (Women_business_and_law)	0.539773
Governance	Zscore (Gov_effectiveness)	0.200258
	Zscore (Regulatory_quality)	0.205952
	Zscore (Voice_and_account)	0.178378
	Zscore (Rule_of_law)	0.209919
	Zscore (Control_of_corruption)	0.205493

Source: calculated by the authors based on the World Bank (n.d.-a); ILO (n.d.); World Inequality Lab (n.d.); and World Bank (n.d.-b).

). For the living conditions and equality pillars, two or three factors were retained, and the variance of each factor was considered when determining the weights of each component in the pillar.

The rank of developing countries was obtained by aggregating all pillars equally into the overall inclusive growth index (

Table 7. The rank of developing countries by the inclusive growth index (IGI).

Country	IGI	Rank
Lithuania	0.908304	1
Latvia	0.862418	2
Croatia	0.845399	3
Poland	0.815262	4
Hungary	0.80783	5
Uruguay	0.787841	6
Costa Rica	0.759116	7
Romania	0.685051	8
Serbia	0.673443	9
Chile	0.670681	10
North Macedonia	0.669883	11
Bulgaria	0.665943	12
Malaysia	0.665315	13
Argentina	0.650599	14
Panama	0.649071	15
Albania	0.647577	16
Moldova	0.614854	17
Georgia	0.601012	18
Armenia	0.598809	19
Thailand	0.594246	20
Kazakhstan	0.591266	21
Dominican Republic	0.587873	22
China	0.572039	23
Turkiye	0.557419	24
Azerbaijan	0.551969	25
Paraguay	0.526683	26
Colombia	0.516751	27
El Salvador	0.515984	28
Indonesia	0.511156	29
Philippines	0.508721	30
Peru	0.507188	31
Morocco	0.497556	32
Mexico	0.496227	33
Mongolia	0.494077	34
Sri Lanka	0.472266	35
Brazil	0.469085	36
Nepal	0.451617	37
Bolivia	0.431097	38
Senegal	0.429195	39
Jordan	0.427443	40
Bangladesh	0.423656	41
Ghana	0.418823	42
Nicaragua	0.418409	43
Russian Federation	0.407606	44
India	0.406232	45
Algeria	0.401814	46
Tunisia	0.400728	47
Kyrgyz Republic	0.396926	48
Tanzania	0.396055	49
Rwanda	0.39448	50
Egypt, Arab Rep.	0.390041	51
Namibia	0.384161	52
Guatemala	0.36708	53
Honduras	0.364681	54
Kenya	0.349192	55
Mauritania	0.343504	56
South Africa	0.330583	57
Tajikistan	0.317745	58
Lesotho	0.30105	59
Uganda	0.299873	60
Sierra Leone	0.296784	61
Nigeria	0.293112	62
Pakistan	0.29243	63
Lao PDR	0.287722	64
Mali	0.284949	65
Cameroon	0.272625	66
Iran, Islamic Rep.	0.264038	67
Zambia	0.261656	68
Madagascar	0.249335	69
Burundi	0.240851	70
Mozambique	0.235461	71
Zimbabwe	0.209607	72
Chad	0.18125	73

Source: calculated by the authors based on the World Bank (n.d.-a); ILO (n.d.); World Inequality Lab (n.d.); and World Bank (n.d.-b).

). Developing countries with better economic performance achieved higher levels of inclusive growth. Lithuania and Latvia were the highest-ranked developing countries, leading in four pillars, whereas Chad, Zimbabwe, and Mozambique were the lowest-ranked. The construction of a composite index of inclusive growth revealed disparities between developing countries. Developing countries were a heterogeneous group, ranging from 0.181 to 0.908. The pillars provided some insight into the main reasons for these differences. According to the results, the highest level of heterogeneity was demonstrated in the economic development and governance pillars. In this context, developing countries with better economic performance and governance indicators were highly ranked, whereas the least developed countries with a lack of effective governance were ranked lowest. Simultaneously, developing countries mostly underperformed in the equality pillar, emphasizing the urgent need to address inequality in these countries. This suggests that inequality becomes more prominent at a certain level of economic development and governance.

To illustrate the explanatory capacity of the inclusive growth index, its linkages to well-known indices were examined. Figure 1 shows that the inclusive growth index was an optimal measure in the context of its high correlation with the SDG index, EBRD inclusion score of the “assessment of transition qualities”, HDI, and Social Progress Index. This shows that the suggested inclusive growth index aligns with human development, social progress, and sustainable development goals. In other words, good performance in the inclusive growth index indicates progress in the above-mentioned measures and vice versa.

The results of the first part of the econometric analysis of the relationship between the inclusive growth index and changes in sectoral composition of the economy are presented in

Table 8. OLS regression results for Equations (3)–(5) (dependent variable: Log_IGI).

Explanatory Variables	Model 1	Model 2	Model 3
Log_VAA	−0.308387 ***	−0.205082 *
	(0.0455)	(0.102366)
Log_VAI	−0.196318		0.15391
	(0.172874)		(0.13998)
Log_VAS		0.535469	1.19562 ***
		(0.418662)	(0.21199)
Control Variables
Log_CPI	−0.140236	−0.140909	−0.15674 *
	(0.084960)	(0.087353)	(0.07335)
Log_GFCF	0.126965 (0.140113)	0.135480 (0.149422)	0.18897 (0.13410)
Number of Observations	64	64	64
Adjusted R2	0.4516	0.4659	0.3904

Figures in parentheses show robust standard errors for models 1 and 2, standard errors for model 3, * 0.01 < p ≤ 0.05, and *** p ≤ 0.001. Source: calculated by the authors based on R programming.

.

For all three models, Shapiro–Wilk normality and studentized Breusch–Pagan tests were run, and the results showed that model 1 met the homoskedasticity requirement, but the residuals of the model were not normally distributed, and the residuals of the second model had non-constant variance and were not normally distributed. Since key OLS assumptions were violated for the first two models, robust standard errors were used for them. The third model parameters met the homoskedasticity and normal distribution requirements, and standard OLS was employed for the model.

The results of the first model showed that the coefficient of VAA was statistically significant and negatively related to the inclusive growth index. This indicated that the agriculture–service shift would have a positive impact on the inclusive growth of developing countries. Specifically, when the share of industry was constant, a decline in the share of agriculture and an increase in the share of service would accelerate inclusive growth in developing countries.

The results of the second regression indicated that the coefficient of VAA was statistically significant and negative, but the same was not true for VAS. When the share of service was constant, a decline in the share of agriculture and an increase in the share of industry would increase the inclusive growth index, but the service–industry transition would have no impact on the inclusiveness of developing economies.

In the third model, we can see that the coefficient of VAS was statistically significant and positive. One more time, the results of the model showed that the agriculture–service transition would boost the inclusive growth of developing countries. The third model showed that the consumer price index could have a negative impact on the inclusive growth of the economy. This was mainly through its distributional effects on income and poverty.

In summary, the results of our analysis show that agriculture–service and agriculture–industry transitions would increase the inclusive growth of developing economies.

The results of the first part of the econometric analysis of the relationship between the inclusive growth index and changes in the sectoral composition of the economy are presented in

Table 9. OLS regression results for Equations (6)–(8) (dependent variable: Log_IGI).

Explanatory Variables	Model 4	Model 5	Model 6
Log_EA	−0.27458 *** (0.05793)	−0.22596 ** (0.07636)
Log_EI	0.07534 (0.10121)		0.11309 (0.10897)
Log_ES		0.18880 (0.15423)	0.47771 *** (0.12599)
Control Variables
Log_CPI	−0.18115 ** (0.06668)	−0.16946 * (0.06567)	−0.18340 * (0.07058)
Log_GFCF	0.05689 (0.11635)	0.08277 (0.11383)	0.08224 (0.12369)
Number of Observations	64	64	64
Adjusted R2	0.4938	0.501	0.438

Figures in parentheses show standard errors, * 0.01 < p ≤ 0.05, ** 0.001 < p ≤ 0.01, and *** p ≤ 0.001. Source: calculated by the authors based on R programming.

.

For all three models, Shapiro–Wilk normality and studentized Breusch–Pagan tests were run, and the results showed that they met the homoskedasticity and constant variance requirements. Standard OLS was applied to the models.

The results of model 4 showed that the coefficient of EA was statistically significant and negatively affected inclusive growth. This indicated that when the share of employment in industry was constant the shift of employment from the agricultural to the service sector would have a positive impact on the inclusive growth of developing countries. But the impact of the industry–service shift was not significant.

The results of model 5 indicated that the coefficient of EA was statistically significant and negative. When the share of employment in the service sector was constant a decline in the share of employment in agriculture and an increase in the share of employment in industry would increase the inclusiveness index, but the service–industry employment shift would have no significant impact on the inclusive growth of developing economies.

In model 6, we can see that the coefficient of ES was highly significant and positive. It proved the assumption that when the share of employment in industry was constant an increase in the share of employment in the service sector and a decline in the share of employment in agriculture would positively affect the inclusive growth of developing countries.

All three models again verified the assumption that the consumer price index has a negative distributional impact on the inclusive growth of the economy.

In summary, the results of the second part of the regression were compatible with the previous results, proving that the transformation of the economy, particularly the shift from the non-productive agricultural sector to the productive service sector, would accelerate the inclusive growth of developing economies.

5. Discussion and Policy Implications

The assessment of inclusiveness through the composite index of inclusive growth is an inseparable part of the development and implementation of national strategies toward inclusive growth. The basis of all indices is the same logic: the areas of inclusiveness are distinguished and then the indicators are chosen that describe those areas (aspects) of inclusive growth, selecting the methodology that is used to unify those indicators into one composite index.

However, there is no generally recognized methodology or precise measure of inclusive growth. All indices have limitations regarding the indicators included in the pillars and the methodologies applied for integrating those indicators into one composite index. The first issue is related to how comprehensively the indices measure inclusiveness. From this perspective, the vast majority of indices evaluate particular areas of inclusiveness, excluding some aspects of inclusiveness in the scope of the assessment. Therefore, the set of indicators used in this study attempted to cover all areas of inclusiveness. For this purpose, economy, living conditions, and equality pillars and a range of indicators were broadly used by UNCDAT and the World Economic Forum; however, considering the crucial role of effective governance in development, similar to the OECD and Asian Development Bank, we added the pillar of governance, containing world governance indicators (the rule of law, government effectiveness, voice and accountability, regulatory quality, and the control of corruption). Another problem is related to the different number of indicators included in the pillars that can have an impact on indicators’ significance and participation in the aggregating index. However, this issue arises mostly when pillar weights are obtained through PCA, whereas, in our methodology, when pillars were weighed equally, this was less problematic.

There is also a limitation related to poor country coverage of indicators, and some indicators were not possible to include in the inclusive growth index as PCA requires a large number of countries with full indicator coverage. Therefore, separate pillars were not created, and indicators were distributed among other pillars. For example, as frequently observed in the literature, we included environmental indicators into the second pillar, living conditions, and, along with adjusted net savings, generated the factor environmental quality. It should be noted that the limitation of data coverage also emerged while assessing the impact of determinants on the inclusive growth index. Hence, the indicator frameworks may later require some corrections as the need to include new indicators and dimensions may arise as new indicators become available.

The most serious methodological issue is related to attributing weights to variables. We used PCA-based factor analysis to reduce dimensionality and derive data-driven factor loadings to define the weights of original indicators in the pillars. This approach was effective in eliminating noise and captured the most essential patterns (Broby & Smyth, 2025). This helped to consider each indicator’s weight, determined by its comparative significance. Although applying the PCA approach helped to avoid bias and subjectivity, there was another problem related to stability and volatility. To partially solve the issue of huge fluctuations, we aggregated the pillars into the inclusive growth index equally. Moreover, equal weights for pillars were also justified considering the fact that the application of PCA for pillars was not appropriate for developing countries as, while employing PCA, the pillar equality had the lowest component loading, which meant that it had the lowest correlation with other economic indicators and weighed the lowest among other pillars. This posed a challenge for interpretation and contradicted the essence of inclusive growth based on the concept of providing economic growth that is distributed to everyone equally. Therefore, to solve this issue, all four pillars were weighed equally.

In addition, the composite index needed to be tested for its robustness. This definitely helped to provide some measurement of the influence of the selection of original indicators, normalization of data, weighting, and aggregation. A sensitive exercise was conducted comparing the ranking scores of the inclusive growth index with equal weights of pillars, with the ranking score of IGI weights obtained using PCA (Figure 2). While comparing these two alternative approaches, a slight difference was noticed whereby no country showed the maximum of a 0.1 ranking score difference. It is obvious that a change in weights did not have a significant impact. Therefore, this could not be interpreted as a methodological issue as many countries demonstrated very similar scores.

However, it was still not enough to exclude subjectivity and provide stability for the assessment. The change in the indicators and the list of countries impacted the results of the PCA. Therefore, systematic corrections and adjustments needed to be made to improve the composite index of inclusive growth.

The analysis of the nexus between inclusive growth and economic transformation and the assessment of the influence of structural changes on the inclusive growth index were critical to understanding the distributional effects of the structural transformation of developing economies. Since previous research generally encompasses the impact of structural transformation on income distribution and poverty, there is no well-adapted and recognized method for assessing the above-mentioned relationship.

Nevertheless, the approach developed by the authors has some limitations. The first issue is related to the limits of the data range due to many missing observations for the year 2023 in government expenditures on education, R&D, social protection, and social insurance, which were among the main drivers of the inclusive growth of the economy. For this, we discussed different combinations of control variables such as the above-mentioned expenditures, general government final consumption expenditures, households’ final consumption expenditures, gross fixed capital formation, trade openness, individual internet users’ share, share of rural population, inflation rates, and many others not calculated in the estimation of the inclusiveness index but having considerable impact on the inclusive growth of the economy. The results showed that the discussed relationships were statistically significant and theoretically proven, with gross fixed capital formation and consumer price index being considered as control variables to reduce individual country-level differences. Since the main data in our regressions came from the World Bank database and the source of most of the statistics was the statistical bodies of member countries, there could have been discrepancies in data quality across the observed countries. The lack of data and poor quality of the relevant data drove the authors to apply cross-sectional analysis for a single point in time, which made it harder to control unobserved heterogeneity and did not show changes over time or the direction of the effects. So, to decrease the overall error bias, robust standard errors were estimated for the first two regressions. This approach was more resistant to heteroskedasticity and non-normal distribution issues.

The next issue was related to the insignificance of gross fixed capital formation in the analysis through its potential impact on the inclusive growth of the economy. It is a well-known fact that capital formation plays a significant role in job creation, infrastructure development, and productivity and wages, enabling access to public goods and services in developing countries. Therefore, this factor might have had a key role in boosting inclusive growth.

Nevertheless, the overall results of both parts of the regression were consistent with the widespread assumption that the path to inclusive growth for developing countries is the shift from agriculture to services, which is also proven in practice.

6. Conclusions

Over the last few decades, inclusive growth has been in the scope of interest of both economists and international organizations. It is broadly recognized that people from different social groups take part in the growth process and reap the benefits of growth equally. The concept of inclusive growth is based on equal contributions to the improvement of everyone’s life quality. In this context, it is vital to measure countries’ progress in social inclusiveness to provide guidance for policymakers to identify gaps and adjust and devise effective policy mechanisms.

However, the evaluation of inclusive growth is challenging due to the multifaceted nature of inclusiveness and the lack of universally conceptualized, measured characteristics and approaches to measuring inclusive growth. Examining the indicators and methodologies, we suggested an alternative methodology like UNCDAT that applies PCA within pillars to obtain weights for original variables within each pillar, avoiding subjectivity and bias, and the attainment of equal weights for pillars to avoid year-on-year volatility. Based on this methodology, we evaluated the inclusive growth of 73 developing countries and revealed homogeneity in performance regarding inclusive growth. For this purpose, four pillars of inclusive growth were distinguished: economy, living conditions, equality, and governance. It should be noted that international organizations including UNCDAT mostly use the first three pillars while assessing inclusive growth. However, like the OECD and Asian Development Bank, we added pillar four, governance, since the progress and process of inclusive growth in developing countries is highly related to the level of development of public institutions and regulators. The inclusive growth index demonstrated explanatory power, making it an efficient analytical measure, considering its linkages with other relevant indices of inclusiveness, sustainability, social progress, and human capital.

According to the results, the group of developing countries behaved more heterogeneously, mostly due to the level of economic development and governance effectiveness. Hence, the countries (Lithuania and Latvia) that succeeded in both led in the ranking, and their indicators could be considered as targets for others. In the least developed countries, essential measures were required to be initiated toward inclusive growth. It is concerning that developing countries exhibited weaker outcomes regarding equality, and the economic development indicators were less correlated with equality indicators, highlighting the need for active reforms in this direction. In this context, the composite inclusive growth index can be an effective tool for assessing countries’ progress in inclusive growth, benchmarking countries, and revealing the main disparities between countries. It can also play a vital role while adjusting priorities, reviewing targets, and making recommendations in strategies toward inclusive growth.

Structural change in the economy is vital for developing countries in the context of promoting decent and full employment, enhancing productivity, increasing investments in infrastructure, and ensuring access to public goods and services. The path of structural changes in developing countries is the shift from the agricultural to the service sector. According to the results of the regression analysis, service sector growth is the driver of the inclusive growth of developing economies. This can be explained by the fact that the shift to the service sector can boost overall labor productivity and increase workers’ earnings. It is worth mentioning that according to the results of the regression model, if the share of the service sector is constant, the main driver of inclusive growth becomes the industry sector, and the agriculture–industry shift can positively affect growth.