Roberta Barbieri, Francesco Natale, Valentina Cattivelli (p), Benedetta Coluccia (p)

Ashley Burdett

see long abstract

Ashley Burdett (p), Matteo Richiardi

Dawid Majcherek

TThis paper examines the impact of population ageing on digital skills in four European economies (Poland, Hungary, Italy, Greece) characterised by economic models and welfare regimes, as well as exhibiting different population trajectories. We explore this relationship under a number of different assumptions regarding how AI will impact the composition of the labour market.

Europe, like many other advanced economies, is experiencing a period of significant demographic change. An extended period of low fertility and increasing life expectancy, is resulting in rising old-age dependency ratios and will eventually mean shrinking populations. Since the 1970s, fertility rates have been declining across EU countries, with rate below the replacement rate since 1995 (Bignami et al 2024). Simultaneously life expectancy has been increasing steadily in Western Europe, but most rapidly among Eastern European countries (Hrzic and Vogt 2024, Hrzic et al. 2020). Whilst net migration has helped to offset this trend (Grundy and Murphy 2017), from 2026 the population of the EU is expected to decline (Eurostat 2023).

There is a large literature highlighting that population ageing will exert considerable pressure on public finances by affecting labour markets, capital accumulation and growth, the composition of demand and balance of payments as well as public expenditures and incomes (Börsch‐Supan 2003, Börsch‐Supan et al. 2006, Krueger and Ludwig 2007, Attanasio et al 2007, Bodner and Nerlich 2022, Ludwig and Vogel 2014, Lisack et al 2019, Maestas et al 2023 ). Most directly, public spending will increase as the number of pensioners requiring pension payouts and health and long-term care rises, whilst the working age population shrinks. The latest EU projections suggest that additional costs of ageing for government will increase in all EU countries reaching an average of 25.6% of GDP in 2070. (European Commission 2024).

One potential means of offsetting this fiscal pressure is to boost productivity, thereby expanding the economic base and easing public financing. Europe, however, has been experiencing a period of stagnated growth primarily driven by the lack of tech businesses and technology diffusion. Overall, labour productivity in Europe is currently below 80% of that in the US, however omitting tech sectors, EU productivity gains have been in line with those of the US. (Draghi 2024) EU policymakers, therefore have set a growth focussed agenda centred in part on the “Digital transition”. Thus, through initiatives such as the Digital Decade Policy Programme and the Recover and Resilience Facility, European policymakers have set out a large agenda to harness the next technological revolution (AI), to facilitate competitiveness and generate future growth.

A key aspect of this agenda is the broad development of digital skills across the continent. Numerous programs and funds have been set up with the explicit task of supporting digital skills . Digital literacy will not only nurture tech businesses will generate innovation, but it will also be important to enable existing businesses in traditional industries to integrate new technologies into their operations to improve efficiency thus ensuring they can remain competitive (Ritz et al 2019, Oduro et al 2024). Furthermore, Europe’s commitment to inclusive growth requires widespread digital literacy to ensure all can participate in the digital economy and enable engagement with increasingly digitised public services. (Draghi 2024). In addition, there is evidence that older population having good digital skills may increase the longevity work-lives (Bartel and Nachum 1993, Friedberg 2003, Sanzenbacher 2025) and facilitate provision of care and health services which may free up human resources for more productive activities.

Of course, a potential obstacle to this solution is the nature of the demographic transition itself. Older individuals on average have fewer digital skills (Farooq and Suomi 2022) and typically find in more challenging adopt new skills (Blazic and Borak 2018). This may potentially undermine the digital transition as the share of the elderly in population grows. In this paper we investigate the existence and extent of this phenomenon.

Focussing on four European countries at different stages of their demographic change and on different trajectories – Italy, Greece, Poland and Hungary, to this aim we utilize population projections from Eurostat and an individual measures of digital skills developed by Richiardi et al (2024) . Their index is constructed using an Item Response Theory (IRT) model applied to data from the Community Survey on ICT usage in households and by individuals (2015-2016,2017, 2019). The IRT model aggregates information from numerous items (dummy variables) into a single continuous index an underlying latent trait, here digital skills, accounting for each items difficulty and ability to discriminate among individuals with similar levels of digital skills. To extend the use of this measure to other datasets, the authors regress the digital skills index on a number of demographic variables to generate a predictive model. We utilise these estimates to obtain a measure of digital skills for our representative sample for each of our countries of interest.

To obtain data for the determinants of digital skills from 2011 to 2040 we utilize projections from SimPaths, a dynamic microsimulation model (Bronka et al. 2025). Although originally constructed for the UK, we have adapted and calibrated the framework for each of our sample countries. SimPaths projects forward simulated trajectories of a cross-sectional population sample with respect to four main interrelated life domains: family, work, education and health based on some biological, institutional or behavioural rules. The model aims at tracking the complex pathways between individual life domains.

A primary advantage of this approach is its ability to act as aggregator of countervailing forces. Every force in the model pushes in a potentially different direction, some reinforcing, others counterbalancing. SimPaths operates as a collector of all these forces, and a calculator of their interaction and composition. For instance, the effects of population ageing on the labour force can be counteracted by increasing female labour force participation, although the need to provide care to elderly relatives might attenuate it. Unlike standard reweighting techniques, SimPaths endogenously models these behavioural interactions, providing a more nuanced projection. We configure the model for each country using Eurostat data and longitudinal EU-SILC to set initial conditions and transition parameters.

With a representative, projected measure of individual digital skills for our sample countries from 2012 to 2040 we analyse the data to assess the role of population ageing. Highlighting two key results:

First, we isolate the impact of population ageing by comparing the projected series of the population’s mean level of digital skills against a counterfactual scenario in which the 2012 age structure is held constant. We find that for each case population ageing significantly reduces the average level of digital skills.
Interestingly, countries that the most severe predicted population ageing, Italy and Greece, exhibit significant gains, although ageing is non-negligible. Decomposing these trends reveals that favourable underlying dynamic mute the negative impact of ageing in these countries. In Italy, a relatively rapidly rising share of the adult population that has a tertiary education counterbalance the ageing effect. In Greece, the model suggests that a post-crisis economic recovery protects digital skills.

Second, we explore a channel through which the digital transition itself might alter our projection: occupational structural change due to AI. When imputing our measure of digital skills, we make assumptions about the occupational structure of each country. We explore the impact of technological developments increasing the share of non-manual occupations (if AI is complementary to cognitive skills) and, on the contrary, increasing the share of manual occupations increases (if AI is a substitute for cognitive skills). Our projections indicate that digital skills in all countries would increase if a greater share of occupations increases become non-manual. Greece is the most sensitive to declines in the share of non-manual occupations, whilst Poland is the least so. However overall impact of ageing is an order of magnitude larger.

Dawid Majcherek (p), Arkadiusz Kowalski, Małgorzata Lewandowska, Desislava Dikova, Scott Hegerty

Valentina Cattivelli

The digitalization of healthcare presents a significant opportunity to reduce inequalities in access to medical services across the European Union. The rapid adoption of digital health technologies during the COVID-19 pandemic demonstrated their potential to mitigate challenges in healthcare delivery. Building on this momentum, we compiled a dataset for the 27 EU countries using the European Health Interview Survey (EHIS), the Digital Economy and Society Index (DESI), and other Eurostat sources. Applying k-means cluster analysis, we categorized EU countries based on two key dimensions: digitalization levels and disparities in healthcare access. Our findings reveal five distinct clusters, with two clusters experiencing high, two facing moderate, and one showing low levels of unmet medical needs. Notably, only the Nordic countries, Spain, and Cyprus demonstrate high digital readiness. Meanwhile, Western European nations with strong healthcare systems fall into a cluster characterized by moderate levels of both digitalization and unmet need. The majority of EU countries still require substantial investment in digital infrastructure to expand the adoption of digital healthcare solutions among patients and professionals. To enhance equitable healthcare access, policymakers must prioritize initiatives that foster digital health innovations as an essential component of modern healthcare systems.