Who Gets Left Behind Online? New Study Maps the Resources Behind Older Adults’ Technology Use

As banking, healthcare, government services, and social life migrate online at breakneck speed, a new study from South Korea offers one of the most detailed portraits yet of why some older adults thrive in the digital world while others are left staring at a locked screen. The research, published in BMC Public Health by Jiyeon Ha of Ajou University, Joohyun Chung of the University of Massachusetts Amherst, and Juah Kim of Korean Bible University, analyzed data from 8,664 older Koreans and found that technology use in later life is not a simple matter of access or willingness. Instead, it rests on a layered stack of resources—economic, social, cognitive, and physical—that either accumulate to enable digital engagement or quietly conspire to block it.

The team grounded their investigation in Resources and Appropriation Theory, a framework that treats technology adoption not as a single decision but as the end point of a chain of resource conversions. In this view, having money is not enough; money must translate into devices and connectivity, which must in turn be converted into skills and confidence through social support, and finally into actual use, which depends on the cognitive and physical capacity to operate the technology. Each stage of that conversion can fail, and each failure compounds the one before it. By testing resources in successive blocks, the researchers could see which layers mattered most once the others were held constant—a statistical unpacking of the digital divide that goes far beyond the crude question of who owns a smartphone.

The data came from the 2020 National Survey of Older Koreans, a nationally representative survey approved by the institutional review board of the Korea Institute for Health and Social Affairs. The researchers excluded people who used no information and communication technology devices at all and those whose interviews had been completed by a proxy respondent, leaving an analytic sample of 8,664 adults. ICT utilization was measured with eleven binary items covering different devices and activities, combined into a single score ranging from zero to eleven. A tetrachoric exploratory factor analysis, reported in supplementary materials, confirmed that the eleven items shared a single underlying dimension, meaning the score could reasonably be treated as one coherent measure of digital engagement rather than a grab bag of unrelated behaviors.

The statistical architecture of the study was deliberately incremental. Model 1 included personal and positional categorical inequalities such as age, sex, education, and employment. Model 2 added material resources in the form of total annual personal income. Model 3 added social resources, captured through assistance with daily activities. Model 4, the fully adjusted final model, added cognitive and health resources. All analyses were survey-weighted hierarchical linear regressions, and the team checked robustness using alternative model specifications, including different income definitions, count-model approaches such as Poisson and negative binomial regression, and a model adding the number of chronic diseases. This kind of sensitivity testing matters, because in observational research the conclusions should not hinge on one arbitrary analytical choice.

The headline findings were strikingly consistent. In the final model, older age was associated with lower ICT utilization, while higher educational attainment, personal income above 120 percent of the sample median, the absence of subjective cognitive decline, and better subjective health status were all associated with higher utilization. The final model explained 43.2 percent of the total variance in ICT use—an unusually large share for research on technology adoption, suggesting that these resource-based factors capture much of what really drives digital engagement in later life. Education, in particular, emerged as a powerful predictor, consistent with the idea that schooling builds not just knowledge but a durable capacity for learning new systems, interfaces, and procedures decades after graduation.

Not every result was straightforward, and the study is refreshingly honest about where the evidence wobbled. Middle-income status, defined as 80 to 120 percent of the median, showed lower ICT utilization than the lowest income bracket in the primary model—a counterintuitive result that vanished when the researchers used equivalized household income instead of personal income. This suggests the apparent middle-income disadvantage may have been an artifact of how income was measured, perhaps reflecting households where resources are pooled and shared rather than held individually. The lesson for future research is methodological as much as substantive: how you operationalize economic resources can change the story you tell about inequality.

Two other variables proved sensitive to model specification in ways that reveal the complexity of the divide. Employment was significant in the primary model but lost significance in two of the three alternative specifications, hinting that any apparent employment effect may be entangled with age, health, or income rather than being an independent driver of technology use. Sex, by contrast, was non-significant in the primary model but significant in all three alternative specifications, a pattern the authors flag as unstable. Taken together, these results caution against simple narratives—such as the idea that older men or working seniors are inherently more digital—and instead point to a web of overlapping disadvantages that statistical models must carefully disentangle.

The public health stakes of this work are considerable. The authors frame the rapid integration of ICT into daily life as a force that threatens to widen both the digital divide and health inequalities among older adults. When appointment booking, prescription refills, telemedicine, and public health information move online, those who cannot use these tools face more than inconvenience; they face systematic exclusion from services that determine health outcomes. The COVID-19 pandemic accelerated this shift dramatically, making digital access a de facto social determinant of health. A study identifying cognitive decline and poor self-rated health as independent barriers suggests that the people most in need of health services may be the least equipped to reach them digitally—a cruel paradox that policy must actively correct.

The authors conclude that ICT utilization among older adults is associated with sociodemographic, economic, cognitive, and health-related factors, and they call for tailored interventions that consider individual characteristics rather than one-size-fits-all digital literacy campaigns. In practice, that could mean programs that pair technology training with cognitive support for people experiencing decline, income-sensitive subsidy schemes that recognize the gradient in digital capacity, and outreach designed for those with lower educational attainment. Because the study is a cross-sectional secondary analysis, it identifies associations rather than proving causation, and the authors are careful not to overclaim. Still, with 43.2 percent of variance explained and a nationally representative sample, the findings give policymakers a concrete map of where the digital divide in aging is deepest—and a strong argument that closing it is a matter of health equity, not just convenience.

What makes this study resonate beyond South Korea is the universality of its underlying logic. Rapidly aging societies across East Asia, Europe, and North America are all confronting the same collision between demographic change and digital transformation. The Resources and Appropriation Theory lens offers a transferable insight: interventions that target only one layer—cheap devices, for example, or free Wi-Fi—will fail if the deeper layers of skill, confidence, cognition, and health remain unaddressed. As governments worldwide push citizens toward digital-only services, this research is a timely reminder that the fastest way to widen health inequality is to assume everyone can follow. The slowest adopters are not refusing the digital world; they are running out of the resources needed to enter it, and society can choose to supply them.

Subject of Research: Resource-based factors influencing information and communication technology utilization among older adults in South Korea

Article Title: Aging in a digital world: what influences ICT utilization among older adults?

Article References: Aging in a digital world: what influences ICT utilization among older adults?. (n.d.). https://doi.org/10.1186/s12889-026-29895-5

Image Credits: AI Generated

DOI: 10.1186/s12889-026-29895-5

Keywords: digital divide, older adults, ICT utilization, aging, health inequality, digital literacy, South Korea, public health, socioeconomic factors, cognitive decline, technology adoption, gerontology