Lies, damned lies, and South Korean opinion polls
Opinion polls are big in Korea... To twist the well known adage, there are lies, damned lies, and South Korean opinion polls.
Opinion polls are big in Korea. You can find one to support your views, no matter what they are. Now, this is concerning in itself, and from what I’ve seen, it pays to be pretty skeptical about polls when it comes to gauging public sentiment. Flawed methodology, the failure to take into consideration cultural traits and social desirability bias, compromised data collection practices, as well as institutional bias, should make a researcher think twice about relying on polls. To twist the well known adage, there are lies, damned lies, and South Korean opinion polls.
Before exploring the question of opinion polls, it’s important to recognize that there’s a strong cultural preference for quantitative research in South Korea.
The nation’s focus on measurable outcomes, rapid economic development, a strong American (as opposed to European) influence in academia, and the pressure to publish, means quantitative research dominates. Government-funded projects and academic fields like STEM, public policy, and education prioritize numerical data for its perceived objectivity and applicability to evidence-based decision-making. This trend is reinforced by Korea’s strong emphasis on global competitiveness in research and its data-driven approach to evaluating policies, particularly in areas like education and economic growth.
This is changing slowly. There is a growing interest in qualitative research to explore societal change, cultural dynamics, and human experiences, particularly in areas like sociology, healthcare, and education. Rapid modernization and the onset of aging populations, and shifting social norms have created demand for qualitative studies to provide deeper context into society’s challenges.
Yet, qualitative research still faces funding limitations and perceptions of lower rigor. Indeed, in most subjects at university, instructors teaching qualitative research tell their students that they will come across grading criteria that require data and tables - students will lose ten percent of their grade unless they include data and tables! Many researchers and students actually “create”, or manipulate qualitative data into numerical data and tables in order to fit into assessment criteria.
Many academic institutions in South Korea continue to regard quantitative research as the only credible form of analysis, and totally dismiss the nuanced insights that qualitative methods can offer. This quantitative dominance in turn elevates the status of opinion polls. Need evidence? Let's do an opinion poll!
Any quantitative research is only as good as its adherence to the underlying methodology - and the first major issue with political opinion polls in South Korea is poor methodology. Accurate polling relies on researcher integrity, rigorous sampling, and strong question design. Unfortunately, polls in South Korea often fall short in these areas.
First, researcher integrity faces challenges due to systemic pressures to publish for career advancement, funding-driven conflicts of interest, and inadequate oversight of ethical standards. The competitive environment in academia, think-tanks and government institutes leads to unethical practices, including data fabrication, selective reporting, and plagiarism, while cultural hierarchies discourage junior researchers from challenging misconduct by superiors.
If you’re a junior researcher at a government institute or an intern at a think-tank, fudging research results (sometimes under instruction by your supervisor) is not unheard of.
Second, sampling in political opinion polls often misses population sub-groups - most noticeably, the youth vote. Individuals in their 20s and 30s are significantly less likely to participate in surveys, creating an over-representation of older, more conservative voters that skews data. Geographically, urban areas like Seoul are often over-sampled, while rural regions receive less representation, failing to capture the full regional diversity of opinions. Connected to this, driven by public mistrust, privacy concerns, and survey fatigue, polls in Korea have a significant nonresponse bias. This occurs when the opinions or characteristics of nonrespondents differ systematically from those who do respond. Pollsters attempt to address these imbalances using statistical weighting, which further adds to sampling error.
Third, question formation in South Korea faces challenges due to cultural, linguistic, and methodological factors that can bias responses or lead to misinterpretation. Cultural norms, such as social desirability bias and deference to authority, often lead respondents to provide socially acceptable or "correct" answers. Social desirability bias—the tendency of respondents to answer questions in a manner they believe will be viewed favorably by others—is a particularly critical factor that pollsters in South Korea often overlook. In a society that places high value on group harmony and conformity, individuals are reluctant to express politically unpopular or controversial opinions.
Similarly, linguistic nuances, including issues with translation and the use of honorifics, can affect how questions are perceived and interpreted, while poorly designed questions—such as those that are vague, double-barreled, or leading—further skew responses. Additionally, South Korean cultural preferences for neutral answers on scales obscures strong opinions. Now all of these issues can be overcome by methodological rigor - but who has time for that when it’s all due by the end of the day?
Foreign researchers lean heavily on opinion polls and surveys to gauge public sentiment and societal trends in South Korea. It’s not unusual to see researchers and analysts use them to explain a specific situation without laying a foot in-country. The XXX-institute says this, or the XXX-government research agency says that - unsurprisingly lining up with the political interests of the institute or the government of the day!
Opinion polls are not always as methodologically rigorous as they appear. For foreign researchers, the only answer is to balance that quantitative research with a little more in-country qualitative research or grab an in-country research partner. It's the only way around the dodgy polls.