For Robert Kaplan to echo Japan’s right-wing nationalists by citing Japanese claims that Japanese colonialism doubled the average Korean’s life expectancy (“When North Korea Falls,” October Atlantic) is akin to quoting neo-Nazi claims that the Nazis improved trains and sanitation in occupied Europe. The assertion ignores the larger picture of suffering, degradation, and bestial cruelty inflicted on Koreans by the Japanese, and it insults the memory of the hundreds of thousands of Korean men, women, and children murdered, raped, and tortured by the Japanese colonial authorities.
Furthermore, Kaplan does not mention one of the more fascinating aspects of China’s relationship with both Koreas. Much of today’s Manchuria and part of Siberia once belonged to a powerful Korean kingdom, Koguryo. Had it not been for a military alliance between the Chinese kingdom of Tang and a rival Korean dynasty that ultimately bartered away all that land in exchange for Tang’s help in defeating Koguryo, Manchuria might today still be Korean land.
Manchuria also harbors nearly 2 million ethnic Koreans, and would have even more were it not for China’s brutal policy of deporting North Korean refugees back to North Korea. These facts may explain why China has blocked access to historical sites attesting to Manchuria’s Korean past and rewritten its histories to recast Koguryo as a part of China, despite the fact that even Zhou Enlai acknowledged that Koguryo was a Korean kingdom.
While these details may seem like arcane history to the West, Kaplan himself notes that the United States has a long record of underestimating historical-ethnic disputes. Spurred by new books and films on Koguryo, today’s South Koreans, including the younger military officers, are surprisingly well aware of Korea’s old stake in Manchuria. Should North Korea implode, China will have even more motivation to preempt Korean nationalism by creating a more malleable North Korean puppet state.
The graphics accompanying Richard Florida’s “Where the Brains Are” (October Atlantic) are a textbook example of how to lie with statistics. Side-by-side color-coded maps of the United States show, for each county, the deviation from the national average in the number of college graduates per 100 residents in 1970 and in 2000. If the deviation had been given in percentage points, the graphic might—or might not—give honest visual support to Florida’s contention that graduates have migrated to select cities.
Instead, the deviation has been measured in number of graduates per 100 residents. In the map for 1970, when the national average was eleven per 100, counties are colored the darkest shade of red only if they had nearly double the national average (twenty-one or more), and the palest hue if the number was one per 100 (only 9 percent of the norm) or fewer. In the map for 2000, when the national average had risen to twenty-four, the darkest gradation represents less than 150 percent of the national average (thirty-four or more per 100), and the lightest represents counties that have 58 percent of the average (fourteen or fewer). Not surprisingly, the first map has a nearly uniform healthy tan, while the second seems to show a draining of the nation’s lifeblood to a few tiny spots.
When it comes to misleading readers, a single graphic can do the work of a thousand words.
Susan Williams Professor of Mathematics
University of South Alabama
The editors reply:
Richard Florida wrote the article “Where the Brains Are,” but our staff designed the graphics for the piece. One of the key concepts in the article is the importance of the density of highly skilled people within a geographic region to that region’s economic growth. The more well-educated people, and the tighter they are packed, the faster a region grows. What’s more, the returns on density do not appear to be constant; they seem to be increasing—that is, each skilled worker adds more to a region’s overall growth potential than the last.
This means that showing the difference between cities as a constant proportion, as Susan Williams suggests—saying one city has twice as many graduates as another, for instance, without comparing the absolute density of graduates in each place—would mislead the reader as to the economic gaps between different cities and regions over time. For instance, if one city has five college graduates per 100 residents, and another has ten, the character and growth prospects of the two places are likely to be broadly similar. But two cities with twenty and forty graduates per 100 residents respectively are likely to be fundamentally different places. The percentage gap is the same in both cases—but it’s the absolute gap that matters.
In any case, let us allay any doubts about whether college graduates are indeed flocking to a select number of cities: they are, disproportionately and incontrovertibly, no matter how one chooses to parse the data.