Experimental Method Reveals True South Korean Unification Preferences

In the main, public opinion surveys dealing with Korean unification employ direct questions within observational study designs. They measure public opinion on the necessity of unification, preferred timing for the process, and other related questions.

These designs have inherent limitations in their (in)ability to capture trade-offs among relevant factors and, more seriously, to elicit true preferences on sensitive items. Direct questions that ask respondents whether they support unification are also underspecified and, frankly, vague. Respondents are often asked something like, “What do you think about unification?” with answers choices “unify as quickly as possible,” “wait until the conditions are appropriate,” and so on. We can be sure that unification will have many moving parts if and when it occurs. It, therefore, does a disservice to the complexity of the issues to discuss them this way.

There is, then, a clear methodological gap. Our new study published in the Korean-language journal National Strategy [국가전략] fills this gap by introducing a multidimensional design (a choice-based conjoint) into the domain of unification perception research.¹⁾피터 워드, 스티븐 데니 and 크리스토퍼 그린. "한국인의 통일방식에 대한 숨겨진 선호: 선택형 컨조인트 분석(Choice-based Conjoint Analysis) 결과를 중심으로" 국가전략 30, no.1 (2024) : 137-163. The authors express their gratitude to 황주희 for her editorial assistance and for engaging in constructive dialogue over the article's analysis. Our approach takes components relevant to unification—such as the governance model of a unified Korea, the treatment of North Korean leadership post-unification, employment opportunities for members of the Korean Workers’ Party, and the deployment of forces to the North—and presents them in a multidimensional “unification scenario” that survey respondents evaluate.

The conjoint method yields better insight into the conditional preferences of South Koreans for unification by disaggregating multifaceted scenarios into discrete attributes and randomly varying their assignment to measure their relative importance. It simulates real-world policy preferences much more accurately than past approaches and thus offers policymakers and researchers a better way to understand public opinion on major policies.

Using ten attributes corresponding to different parts of the unification scenario (read them here), a nationally representative survey of 2,009 respondents administered in the fall of 2021 evaluated eight unique scenarios per person. The exercise, as it appeared to respondents, looked similar to the figure below.

The full article presents the findings in context, but we reproduce a sample of them here in English. Taking the marginal means of the attribute levels, which is a descriptive measure of the importance of any given attribute level with the other levels held constant, we obtained the following insights into what motivates public support for unification in South Korea. The marginal means of the baseline model is provided below.

First, the attribute that most strongly influences the support for any unification scenario is the governance model post-unification. South Korean respondents strongly prefer a unification that establishes a liberal democratic regime across the newly unified Korea. This preference underscores the importance of democratic political governance in the eyes of South Koreans. Conversely, the prospect of unification under a "single-party authoritarian" regime, which is meant to proxy a unification scenario under North Korea’s terms, encounters strong opposition. ²⁾There are serious ethical considerations when asking questions about North Korea, especially when it is intended to measure a preference for rule under Pyongyang’s terms. Nevertheless, it is a valid inquiry, a quantity that should be of sufficient interest, and something that can be measured, albeit imperfectly, using the conjoint method. We do not pose the scenario explicitly (i.e., rule under the Korean Workers' Party or Kim Jong-un), but the implication is clear, and the option is embedded within multiple options of the conjoint design, thus decreasing the likelihood of social desirability bias. There is strong antipathy towards authoritarian governance structures, the likes of which would likely define a unification scenario under terms set by Pyongyang.

Interestingly, there is a notable but less pronounced opposition to a confederation option for unification. While we do not go into it in the paper, it is worth pointing out here that, per our subgroup analysis, only conservatives oppose this confederation plan. Progressives do not actively support it, to be sure, but they do not oppose it either. This suggests some openness to a more gradual and less centralized form of unification, providing autonomy and preserving some level of political diversity, though not to the extent of endorsing an autocratic political future.

Second, the study explores the attitudes toward military presence in northern Korea in a post-unification context. South Koreans have a pronounced preference for a military presence involving either joint US-South Korean forces or exclusively the South Korean military. This indicates a desire for security arrangements perceived as friendly and protective of South Korean interests. The findings also reveal equal support for a UN peacekeeping operation, suggesting an openness to international involvement in ensuring peace and stability in the region. Conversely, the option involving a Chinese military presence is as likely to be rejected as the absence of any military presence, indicating reservations about Chinese influence in a unified Korea.

Relatedly, regarding North Korea's nuclear stockpile, the preferences are clear. There is a strong rejection of any management coalition that includes both the US and China, perhaps reflecting concerns about the implications of such a collaboration for South Korean sovereignty and security. The favored options are for the nuclear stockpile to be managed solely by South Korea or through a joint North-South Korean effort. This preference underscores a desire for a resolution to the nuclear issue that is controlled by Koreans, emphasizing a sense of ownership over the peninsula's nuclear security concerns.

Finally, the financial cost of unification is – predictably – a significant concern for South Koreans. The concept of a "unification tax" to fund the process is met with caution. The only scenario that garners a moderate level of support is a modest 1% flat tax, suggesting a willingness to contribute to unification efforts provided the financial burden is not excessive. The proposal of a special 10% tax on the highest earners in South Korea does not significantly increase support, indicating sensitivity to tax increases, even among those more capable of bearing the financial burden.

Other attributes were involved, but these were the ones that had the largest effects on people’s preferences. Our robustness checks using open-text answers to scenario preferences confirm the conjoint-based findings. The issue of the North-South movement post-unification was also a salient topic in the answers, but none of the options moved the needle in favor or against any given scenario; this suggests, perhaps, that the attribute was of note, but the forced-choice answer was not determined by the options provided.

To express the findings more intuitively, we generated the estimated probability of a scenario being preferred at the distribution's minimum, 50th, and maximum points (i.e., the scenario least likely to be accepted, a coin toss, and the most likely). By way of a conclusion, we provide the results below. We can see what South Koreans most desire reflected in the profiles.

The most supported scenario involves a liberal democratic governance structure implemented in a step-by-step fashion over several years (but no more than five) funded by a 1% flat tax. The border remains closed, but only for three years. There is no restriction on South Korean firms' activities, and peace is maintained with a UN peacekeeping presence. The former North Korean leadership is prosecuted at The Hague. Notably, nuclear weapons are managed by (what would soon be former) South Korea.

The least supported scenario includes unification under single-party authoritarian rule; respondents do not support a unified Korea where the Korean Workers Party or some similar arrangement runs the resulting governance structure. Unification occurs immediately and is paid for with a 10% tax on all incomes. South Korean companies are barred from operating in the former North Korea, and there is no post-unification military presence. The former North Korean leadership goes unpunished, and control over North Korean nuclear weapons is left to a joint US-China coalition.