As an indicator for trust Kim and Kim use agreement with the statement “most people can be trusted.” Though it is widely used, this measure faces continuing questions about its validity. Ermisch et al. (2009), for example, find little correlation between it and experimental measures of trust, leading Nannestad (2008) to doubt whether we have learnt anything about trust through its use. Running trust experiments with the survey participants might strengthen the case that is made here.

The essence of Adam Smith's argument is that there is positive feedback from trust to market activity and vice versa. In this way, a commercial society is built. In Kim and Kim (2019), the authors focus on the second part of this relationship. To rule out the possibility that the results are due to reverse causation from trust to market activity, Kim and Kim use an instrumental variables approach. Specifically, they use two instruments: ownership of assets by neighbors hidden from the authorities back in North Korea and the fraction of arid land in the home province. Using these instruments provides support for a causal link between previous buying and selling activity and current levels of trust.

Yet, neither instrument can be viewed as completely solid. The arid land measure is not available at the county level which means that it may act as a proxy for province level effects that also affect trust. With 70% of respondents coming from a single province, one needs to be cautious. Meanwhile, there could be a link between reports of neighbors’ hidden assets and trust that does not run through market activity. For example, an open and trusting person is more likely to win the confidence of others and learn about their secret assets. In a world where assets are supposed to be close to zero, this would lead to a positive correlation. Conversely, being informed about the neighbor's assets might build trust. Relationships of these kinds would mean a violation of the exclusion restriction, so what we are left with at the end of a paper is suggestive rather than conclusive.

Kim and Kim (2019) also report an absence of a link between prior market experience and trust of fellow refugees. It is not clear why this should be the case, but it follows Osborne et al.’s (2018) experimental study of former child soldiers in Uganda in suggesting that life in a totalitarian society may produce patterns of trust that are highly sensitive to the putative partner.

In seeking to put their article in a wider context, Kim and Kim conclude that higher trust may beget more marketization and this will weaken the grip of the authorities, but this seems too optimistic a picture of the consequences of marketization on social control. There are obviously societies that have large market elements (i.e., the People's Republic of China), but where the authorities still have firm control. That said, Kim and Kim (2019) provide a useful insight into the lives of North Korean refugees and how their lives on one side of the border have shaped their world view on the other side.

作为信任度的指标，Kim和Kim使用“大多数人可以信任”的声明。尽管该措施被广泛使用，但其有效性仍面临持续的疑问。Ermisch等。（2009），例如，发现它和信任的实验方法之间几乎没有关联，导致纳内斯塔德（2008）怀疑我们是否通过使用信任中学到了什么。与调查参与者进行信任实验可能会加强此处的案例。

亚当·斯密（Adam Smith）论点的实质是，从信任到市场活动都会产生积极的反馈，反之亦然。这样，就建立了商业社会。在Kim and Kim（2019）中，作者专注于这种关系的第二部分。为了排除结果可能是由于从信任到市场活动的反向因果关系的可能性，Kim和Kim使用了一种工具变量方法。具体来说，他们使用两种手段：北韩当局隐瞒的邻居对资产的所有权，以及本国省份的干旱土地的一部分。使用这些工具为以前的买卖活动与当前的信任水平之间的因果关系提供了支持。

然而，任何一种工具都不能被视为完全可靠。县级没有干旱土地措施，这意味着它可以代替省级影响信任的省级措施。由于70％的受访者来自单个省份，因此需要谨慎。同时，邻居的隐藏资产报告与信任之间可能存在联系，而这种联系并非通过市场活动来实现。例如，一个开放和信任的人更有可能赢得他人的信任并了解他们的秘密资产。在资产应该接近零的世界中，这将导致正相关。相反，被告知邻居的资产可能会建立信任。这类关系意味着违反排除限制，

金和金（2019）还报告说，先前的市场经验和对难民的信任之间没有联系。目前尚不清楚为什么会这样，但它遵循Osborne等人的观点。的（2018）对乌干达前儿童兵的实验研究表明，极权社会的生活可能会产生对假定伴侣非常敏感的信任模式。

在试图将他们的文章放在更广阔的背景下时，金和金得出结论，更高的信任度可能会带来更多的市场化，这将削弱当局的控制力，但这似乎过于乐观了市场化对社会控制的后果。显然，有一些社会具有很大的市场元素（例如，中华人民共和国），但是当局仍然拥有牢牢的控制权。话虽如此，金（Kim）和金（Kim）（2019）对朝鲜难民的生活以及他们在边界一侧的生活如何塑造另一侧的世界观提供了有用的见解。