Changyong Choi and Jesse Lecy have published a fascinating paper, “A Semantic Network Analysis of Changes in North Korea’s Economic Policy” in the journal Governance. In it they subject the titles of all 1,558 articles published in the Journal of Economic Research (Kyongje Yongu) between 1986 and 2009 to semantic network analysis. They interpret the results as reflecting official discourse and policy, not driving it. They analyze four periods: prior to 1994; 1994-2001; 2002-2005; after 2005/2006.
During the first, classically socialist, period, the most central concepts are “socialism” (juche) and “central management.” On the periphery, there is a small cluster of concepts associated with capitalism, which have negative connotations: “crisis,” “conflict,” “monopoly,” “deceitfulness,” and “dominance.” You’ll have to look at the article to see the maps.
During the second, “Arduous March” period, “socialism” is again the most prominent concept, but “Kim Jong-il,” and “kun” or “gun”—the small jurisdictional unit—move to the center. During this period, a more neutral set of terms associated with the market begin to appear: “demand,” “commodity,” “value,” “price,” and “consumption goods.” Choi and Lecy note that term “communism” appeared in the journal titles 12 times in the first period, but only three times in the second period, and disappeared from Kyongje Yongu titles thereafter. Since then, documents have referred to “the completion of socialist society,” and “communism” has been replaced by “Marxist-Leninism” and “previous theories.” “Previous theories”? Ouch.
The network structure of the third, reformative phase (2002-2005), is clearly distinct from the previous two. Four concepts are central: “capitalism,” “military first” (songun), “socialism,” and “profit.” Choi and Lecy mention a wonderful bit of historical engineering: in order to justify the discussion of pseudo-market policies, the argument was brought forward that in the 17th and 18th centuries, Korea had begun to develop capitalism that was then disrupted by Japanese colonialism.
The final period reveals a kind of dialectical synthesis. The map of the fourth period most closely resembles that of the first. There are five leading concepts: “improvement of people’s life,” “economic development,” “economic strong state,” “People’s economy” (Inmin Kyongje), and “socialism.” “Capitalism” and associated concepts fade in significance. Yet Choi and Lecy observe that in 2006, the journal also introduced a new section titled “Common Knowledge” (Sangsik) which contains relatively neutral discussions of features of the global market economy such as “currency transactions,” “dumping,” “balance of international payments,” “tariffs,” “credit crisis,” “foreign exchange markets,” and “government bond market.”
The authors mention that they did sensitivity analysis on this periodization but I did not see any results. What would have been interesting to see if whether one could use shifts in the semantic network to actually define the periods endogenously—to use these shifts as signals of policy change, rather than defining the policy regimes exogenously and then analyzing for content. Maybe that will be their next paper.