A Cross-Country Econometric Analysis of Digital Access and Income Inequality (2000–2022) This study analyses how digitalisation affects income inequality, through a cross- country econometric analysis covering 27 countries over the period 2000–2022. Its main contribution to the literature is to fill a gap in many previous studies, which focus mainly on specific cases or regional contexts, by offering instead a comparative, longitudinal and systematic analysis of the interaction between digital access and income distribution. The analysis considers the impact of three key dimensions of digitalisation—Internet access, fixed broadband subscriptions and mobile phone services—on inequality, measured through the Gini index and the income shares held by different percentiles of the population. Countries were selected based on three main criteria: availability and completeness of data over the period considered, geographical and development diversity to ensure a solid comparative basis, and minimal presence of digital infrastructure at the beginning of the period. The 2000–2022 time frame was chosen to capture not only the initial phase of digital expansion associated to the new millennium, but also the recent acceleration linked to both the diffusion of broadband and mobile devices as well as Covid outbreak, thus allowing an analysis of medium-long term trends. The data used come from recognized and methodologically harmonized international sources (World Bank, ITU, OECD), and were subjected to consistency checks to minimize regional discrepancies in measurement. From a methodological viewpoint, the study uses Generalized Least Squares (GLS) regression models, corrected for heteroskedasticity and autocorrelation in panel data, combined with random effects models to account for unobservable heterogeneity between countries. Control variables such as Gross Domestic Product (GDP) per capita growth, demographic structure (young and working age population), urbanization rate and other economic-structural factors are included. The analysis was accompanied by extensive diagnostic tests (including VIF for multicollinearity and Wooldridge test), and robustness checks to strengthen the reliability of the estimates. Although the model reduces some risks of bias, it is acknowledged that it does not completely eliminate endogeneity problems, such as reverse causality between inequality and digitalization. The paper tests three main hypotheses, based on the economic and sociological literature: • H1: Greater digital access reduces income inequality by expanding access to employment opportunities, financial services, education and healthcare, especially for low-income groups (Stern, M. J., 2009; George, G. & Schillebeeckx, J. D., 2022); • H2: The redistributive effect of digitalization is mediated by country-specific structural and institutional factors, such as educational systems, redistributive policies and digital literacy levels (Doric, Z., 2022); • H3: In countries where digital access is more widespread and inclusive (e.g. broadband infrastructure and mobile subscriptions), inequality is lower, especially in the lower deciles, indicating that digitalization can foster more equitable economic growth (Afzal, A. et al., 2023). Preliminary results highlight a significant negative correlation between digital access and inequality, particularly in lower-income groups. However, this effect is not uniform: in different national contexts, strong digital divides persist due to infrastructural, economic or cultural constraints, which end up reinforcing existing inequalities. This underlines the dual nature of digitalization: it can be both a tool for inclusion and an amplifier of exclusion, depending on the context and on its usage. The findings have important policy implications. To make digitalization an inclusive lever, public policies must go beyond the mere expansion of access, combining infrastructure investments with targeted initiatives on digital literacy, affordability, and technological endowment. Particular attention must be paid to the most vulnerable groups—such as low-income families or rural communities—who risk being excluded from the digital transition. Digital inclusion must be integrated into broader social and economic strategies to have a concrete redistributive impact. The study offers three main contributions. First, it proposes a comparative and longitudinal econometric model that integrates digital indicators in the analysis of inequalities. Second, it delves into the still little-explored mechanisms that link digitalization and inequality, highlighting the interaction with demographic and institutional variables. Third, it provides empirical evidence useful to support the design of more inclusive digital policies, calibrated to national contexts. The study nevertheless acknowledges some limitations. Despite the applied corrections, potential endogeneity problems remain. Furthermore, the quality of the data, although verified, is uneven across regions and over time. The model does not capture qualitative dynamics such as the role of large platforms or the subjective ability to use technologies. Future research should include micro data and qualitative approaches to explore the underlying behavioral and institutional mechanisms. Insights into intersectional variables—such as gender, age, or urban-rural context— could further enrich the understanding of digital inequalities. In conclusion, this study contributes to the global debate on the role of digitalization as a tool for inclusion or for reinforcing social inequalities. By offering a solid and empirically grounded comparative analysis, it supports the need for proactive digital policies that prioritize not only access, but also skills and the effective inclusion of the most vulnerable segments of the population.
Una analisi econometrica transnazionale dell'accesso digitale e della disuguaglianza di reddito (2000-2022) Questo studio analizza l'impatto della digitalizzazione sulla disuguaglianza di reddito, attraverso un'analisi econometrica transnazionale che copre 27 paesi nel periodo 2000-2022. Il suo principale contributo alla letteratura è quello di colmare una lacuna presente in molti studi precedenti, che si concentrano principalmente su casi specifici o contesti regionali, offrendo invece un'analisi comparativa, longitudinale e sistematica dell'interazione tra accesso digitale e distribuzione del reddito. L'analisi considera l'impatto di tre dimensioni chiave della digitalizzazione – accesso a Internet, abbonamenti a banda larga fissa e servizi di telefonia mobile – sulla disuguaglianza, misurata attraverso l'indice di Gini e le quote di reddito detenute dai diversi percentili della popolazione. I paesi sono stati selezionati in base a tre criteri principali: disponibilità e completezza dei dati nel periodo considerato, diversità geografica e di sviluppo per garantire una solida base comparativa e presenza minima di infrastrutture digitali all'inizio del periodo. L'intervallo temporale 2000-2022 è stato scelto per cogliere non solo la fase iniziale dell'espansione digitale associata al nuovo millennio, ma anche la recente accelerazione legata sia alla diffusione della banda larga e dei dispositivi mobili, sia all'epidemia di Covid, consentendo così un'analisi dei trend di medio-lungo termine. I dati utilizzati provengono da fonti internazionali riconosciute e metodologicamente armonizzate (Banca Mondiale, ITU, OCSE) e sono stati sottoposti a controlli di coerenza per minimizzare le discrepanze regionali nella misurazione. Da un punto di vista metodologico, lo studio utilizza modelli di regressione Generalized Least Squares (GLS), corretti per eteroschedasticità e autocorrelazione nei dati panel, combinati con modelli a effetti casuali per tenere conto dell'eterogeneità non osservabile tra i paesi. Sono incluse variabili di controllo quali la crescita pro capite del Prodotto Interno Lordo (PIL), la struttura demografica (popolazione giovane e in età lavorativa), il tasso di urbanizzazione e altri fattori economico-strutturali. L'analisi è stata accompagnata da ampi test diagnostici (tra cui il VIF per la multicollinearità e il test di Wooldridge) e da controlli di robustezza per rafforzare l'affidabilità delle stime. Sebbene il modello riduca alcuni rischi di distorsione, è riconosciuto che non elimina completamente i problemi di endogeneità, come la causalità inversa tra disuguaglianza e digitalizzazione. Questa tesi verifica tre ipotesi principali, basate sulla letteratura economica e sociologica: • H1: Un maggiore accesso digitale riduce la disuguaglianza di reddito ampliando l'accesso a opportunità di lavoro, servizi finanziari, istruzione e assistenza sanitaria, in particolare per i gruppi a basso reddito (Stern, M. J., 2009; George, G. & Schillebeeckx, J. D., 2022); • H2: L'effetto redistributivo della digitalizzazione è mediato da fattori strutturali e istituzionali specifici di ciascun paese, come i sistemi educativi, le politiche redistributive e i livelli di alfabetizzazione digitale (Doric, Z., 2022); • H3: Nei paesi in cui l'accesso digitale è più diffuso e inclusivo (ad esempio, infrastrutture a banda larga e abbonamenti mobili), la disuguaglianza è minore, soprattutto nei decili inferiori, a indicare che la digitalizzazione può favorire una crescita economica più equa (Afzal, A. et al., 2023). I risultati preliminari evidenziano una significativa correlazione negativa tra accesso digitale e disuguaglianza, in particolare nei gruppi a basso reddito. Tuttavia, questo effetto non è uniforme: in diversi contesti nazionali, persistono forti divari digitali dovuti a vincoli infrastrutturali, economici o culturali, che finiscono per rafforzare le disuguaglianze esistenti. Ciò sottolinea la duplice natura della digitalizzazione: può essere sia uno strumento di inclusione che un amplificatore di esclusione, a seconda del contesto e del suo utilizzo. Emergono anche importanti implicazioni politiche dai risultati. Per rendere la digitalizzazione una leva inclusiva, infatti, le politiche pubbliche devono andare oltre la mera espansione dell'accesso, combinando investimenti infrastrutturali con iniziative mirate su alfabetizzazione digitale, accessibilità economica e dotazione tecnologica. Particolare attenzione deve essere prestata ai gruppi più vulnerabili, come le famiglie a basso reddito o le comunità rurali, che rischiano di essere esclusi dalla transizione digitale. L'inclusione digitale deve essere integrata in strategie sociali ed economiche più ampie per avere un impatto redistributivo concreto. Lo studio offre tre contributi principali. In primo luogo, propone un modello econometrico comparativo e longitudinale che integra gli indicatori digitali nell'analisi delle disuguaglianze. In secondo luogo, approfondisce i meccanismi ancora poco esplorati che collegano digitalizzazione e disuguaglianza, evidenziando l'interazione con variabili demografiche e istituzionali. In terzo luogo, fornisce evidenze empiriche utili a supportare la progettazione di politiche digitali più inclusive, calibrate sui contesti nazionali. Lo studio riconosce tuttavia alcuni limiti. Nonostante le correzioni applicate, permangono potenziali problemi di endogeneità. Inoltre, la qualità dei dati, sebbene verificata, è disomogenea tra le regioni e nel tempo. Il modello non cattura dinamiche qualitative come il ruolo delle grandi piattaforme o la capacità soggettiva di utilizzare le tecnologie. La ricerca futura dovrebbe includere microdati e approcci qualitativi per esplorare i meccanismi comportamentali e istituzionali sottostanti. Approfondimenti su variabili intersezionali – come genere, età o contesto urbano-rurale – potrebbero arricchire ulteriormente la comprensione delle disuguaglianze digitali. In conclusione, questo studio contribuisce al dibattito globale sul ruolo della digitalizzazione come strumento di inclusione o di rafforzamento delle disuguaglianze sociali. Offrendo un'analisi comparativa solida ed empiricamente fondata, sostiene la necessità di politiche digitali proattive che diano priorità non solo all'accesso, ma anche alle competenze e all'inclusione effettiva delle fasce più vulnerabili della popolazione.
The impact of digitalisation on inequalities: a comparative study across nations
SUZZANI, MARTA
2024/2025
Abstract
A Cross-Country Econometric Analysis of Digital Access and Income Inequality (2000–2022) This study analyses how digitalisation affects income inequality, through a cross- country econometric analysis covering 27 countries over the period 2000–2022. Its main contribution to the literature is to fill a gap in many previous studies, which focus mainly on specific cases or regional contexts, by offering instead a comparative, longitudinal and systematic analysis of the interaction between digital access and income distribution. The analysis considers the impact of three key dimensions of digitalisation—Internet access, fixed broadband subscriptions and mobile phone services—on inequality, measured through the Gini index and the income shares held by different percentiles of the population. Countries were selected based on three main criteria: availability and completeness of data over the period considered, geographical and development diversity to ensure a solid comparative basis, and minimal presence of digital infrastructure at the beginning of the period. The 2000–2022 time frame was chosen to capture not only the initial phase of digital expansion associated to the new millennium, but also the recent acceleration linked to both the diffusion of broadband and mobile devices as well as Covid outbreak, thus allowing an analysis of medium-long term trends. The data used come from recognized and methodologically harmonized international sources (World Bank, ITU, OECD), and were subjected to consistency checks to minimize regional discrepancies in measurement. From a methodological viewpoint, the study uses Generalized Least Squares (GLS) regression models, corrected for heteroskedasticity and autocorrelation in panel data, combined with random effects models to account for unobservable heterogeneity between countries. Control variables such as Gross Domestic Product (GDP) per capita growth, demographic structure (young and working age population), urbanization rate and other economic-structural factors are included. The analysis was accompanied by extensive diagnostic tests (including VIF for multicollinearity and Wooldridge test), and robustness checks to strengthen the reliability of the estimates. Although the model reduces some risks of bias, it is acknowledged that it does not completely eliminate endogeneity problems, such as reverse causality between inequality and digitalization. The paper tests three main hypotheses, based on the economic and sociological literature: • H1: Greater digital access reduces income inequality by expanding access to employment opportunities, financial services, education and healthcare, especially for low-income groups (Stern, M. J., 2009; George, G. & Schillebeeckx, J. D., 2022); • H2: The redistributive effect of digitalization is mediated by country-specific structural and institutional factors, such as educational systems, redistributive policies and digital literacy levels (Doric, Z., 2022); • H3: In countries where digital access is more widespread and inclusive (e.g. broadband infrastructure and mobile subscriptions), inequality is lower, especially in the lower deciles, indicating that digitalization can foster more equitable economic growth (Afzal, A. et al., 2023). Preliminary results highlight a significant negative correlation between digital access and inequality, particularly in lower-income groups. However, this effect is not uniform: in different national contexts, strong digital divides persist due to infrastructural, economic or cultural constraints, which end up reinforcing existing inequalities. This underlines the dual nature of digitalization: it can be both a tool for inclusion and an amplifier of exclusion, depending on the context and on its usage. The findings have important policy implications. To make digitalization an inclusive lever, public policies must go beyond the mere expansion of access, combining infrastructure investments with targeted initiatives on digital literacy, affordability, and technological endowment. Particular attention must be paid to the most vulnerable groups—such as low-income families or rural communities—who risk being excluded from the digital transition. Digital inclusion must be integrated into broader social and economic strategies to have a concrete redistributive impact. The study offers three main contributions. First, it proposes a comparative and longitudinal econometric model that integrates digital indicators in the analysis of inequalities. Second, it delves into the still little-explored mechanisms that link digitalization and inequality, highlighting the interaction with demographic and institutional variables. Third, it provides empirical evidence useful to support the design of more inclusive digital policies, calibrated to national contexts. The study nevertheless acknowledges some limitations. Despite the applied corrections, potential endogeneity problems remain. Furthermore, the quality of the data, although verified, is uneven across regions and over time. The model does not capture qualitative dynamics such as the role of large platforms or the subjective ability to use technologies. Future research should include micro data and qualitative approaches to explore the underlying behavioral and institutional mechanisms. Insights into intersectional variables—such as gender, age, or urban-rural context— could further enrich the understanding of digital inequalities. In conclusion, this study contributes to the global debate on the role of digitalization as a tool for inclusion or for reinforcing social inequalities. By offering a solid and empirically grounded comparative analysis, it supports the need for proactive digital policies that prioritize not only access, but also skills and the effective inclusion of the most vulnerable segments of the population.File | Dimensione | Formato | |
---|---|---|---|
2025_07_Suzzani_Tesi_01.pdf
accessibile in internet per tutti
Descrizione: Tesi
Dimensione
4.16 MB
Formato
Adobe PDF
|
4.16 MB | Adobe PDF | Visualizza/Apri |
Suzzani_Executive_Summary_01.pdf
accessibile in internet per tutti
Descrizione: Executive Summary per Controrelatore
Dimensione
1.1 MB
Formato
Adobe PDF
|
1.1 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/239884