The widespread adoption of Information and Communication Technologies (ICTs), or digital technologies, has been uneven, creating digital inequalities rooted in socioeconomic divides. While early research focused on access, the current focus is on the impact of these technologies on societal well-being, moving "beyond the GDP." This paper investigates the relationship between internet use and well-being in Spain, a nation undergoing significant digital transformation. Existing research on well-being has identified various contributing factors such as personal characteristics, circumstances, and contextual influences. Similarly, studies on ICT diffusion consistently show that socioeconomic factors (income, age, education, etc.) shape technology adoption and use. Digital inequalities, stemming from these socioeconomic differences, limit life chances. This study aims to disentangle the complex relationship between internet use and well-being by accounting for the socioeconomic origins of internet use, examining multiple well-being dimensions, and focusing on specific vulnerable groups (elderly, women, disabled individuals). The analysis uses a three-equation model to explain internet use, intensity of use, and the impact of that intensity on well-being, leveraging microdata from the European Social Survey (ESS) for Spain in 2016 and 2018.

Early research on the impact of digital technologies on well-being explored competing hypotheses: the enhancement hypothesis (positive effects through reinforcing offline relationships) and the displacement hypothesis (negative impacts due to substitution of offline activities). Initial findings suggested negative effects, but later studies yielded inconsistent results – some reporting positive, negative, or non-significant links. These inconsistencies are attributed to variations in well-being dimensions examined (life satisfaction, social participation, etc.), technologies analyzed (internet use, social media, etc.), and populations studied. Most studies treat technology use as exogenous, neglecting the influence of socioeconomic characteristics. This study addresses this limitation by explicitly considering the socioeconomic factors driving internet use.

This study utilizes microdata from Rounds 8 and 9 (2016 and 2018) of the European Social Survey (ESS) for Spain, encompassing a sample of 3614 observations. Key variables include five dimensions of subjective well-being (SWB): happiness, life satisfaction, frequency of social meetings, number of people for intimate discussions, and social activity participation. Internet usage is measured by daily internet use (dummy variable) and time spent online (minutes). Socioeconomic variables included age, gender, employment status, partnership status, health status, disability status, income, education level, and residence type (urban/rural). Internet speed data from Fundación Telefónica (2019) was also incorporated. The analysis employs a three-equation simultaneous model: the first equation models daily internet use; the second, the time spent online (conditional on daily use); and the third, the effect of internet usage time on each SWB dimension. The models control for socioeconomic characteristics and internet speed, and include interaction terms for internet time with age, gender, and disability to analyze the effects on specific vulnerable groups. The authors addressed potential convergence problems during model estimation by recoding and treating certain variables as continuous to address this.

The results reveal that the impact of internet use on well-being is nuanced and depends on the well-being dimension and socioeconomic characteristics. Intensive internet use is negatively associated with happiness and life satisfaction, and less frequent social meetings. Conversely, it's positively related to having more people for intimate discussions and greater social activity participation. Regarding vulnerable groups: hampered individuals experienced a stronger negative relationship between internet use and social meetings. No significant gender differences emerged. For older individuals, the negative link between internet time and happiness was attenuated, while the positive association between internet time and social participation was strengthened. Socioeconomic factors also played a crucial role: bad health is negatively associated with all well-being measures; living with a partner is positively related to happiness and life satisfaction but negatively to social meetings and intimate discussions; employment is positively associated with life satisfaction but negatively with social meetings; and income is positively related to all well-being dimensions. Living in a town or small city is positively associated with well-being, while living in rural areas is negatively linked to happiness and social life.

The findings suggest that internet use may substitute face-to-face interactions for online interactions, supporting the replacement hypothesis. This substitution effect appears stronger for those with disabilities. For older adults, online interactions may compensate for reduced face-to-face opportunities. The significant role of socioeconomic factors underscores the need to address digital inequalities. The study’s findings highlight the complexity of achieving well-being improvements through digital technologies, as positive and negative effects may offset each other. Policy interventions should consider the specific dimensions of well-being and the needs of vulnerable groups. The U-shaped relationship between age and well-being was confirmed.

This study contributes to understanding the complex relationship between internet use and well-being. The findings reveal the importance of considering socioeconomic factors and the specific dimensions of well-being when evaluating the impact of internet use. Future research should investigate the types of online activities and their differential effects on well-being, and further explore the heterogeneous impact on different population subgroups. Access to richer datasets, including data from technology companies, could significantly advance this field of study.

The study relies on self-reported subjective measures of well-being, which may be subject to biases. The limited information on internet usage (daily use and time spent) restricts the ability to analyze the effects of specific online activities. The cross-sectional nature of the data limits the ability to establish causal relationships. The reliance on ESS data might lead to some limitations in terms of understanding internet usage patterns and the potential influence of internet speed, requiring additional data sources to support these aspects fully.