The project that I'm currently working on seeks to answer, "why wasn't there any market panic around this whole impeachment thing in South Korea" (despite all the uncertainties)?
I haven't found a very compelling answer that would explicate the empirical result I have, which is one of the reasons why I am presenting it tomorrow at Korean Studies Center (GMU). I hope to garner some ideas as to how to go about this.
It takes a form of public lecture and, to my knowledge, there's going to be no political scientist or economist in the audience. That should explain why first several slides are about super basic stuff.
Thursday, April 27, 2017
Tuesday, April 25, 2017
한국 정치 데이터에서 동성애 보호는 어떤 의미일까?
모두 실시간으로 보지는 못했지만, 4차 대선 후보 토론에서 아마도 소수자 보호에 대한 논란이 조금 있었던 것 같네요. 물론 실제로 건드린 이슈는 동성애 입니다. 동성애는 성적 정체성이므로 합법/불법, 동의/부정의 문제가 아님은 확실합니다.
다만 제 페이스북 타임라인을 보니 많은 문재인 후보가 좀 더 강력하게 동성애자들 (혹은 모든 성적 소수자들)도 일정한 법적 보호를 받아야한다라고 말하지 못한 부분을 아쉬워하는 것 같습니다. 아마도 이런 아쉬움의 기저에는 "표에 눈이 멀어 비겁하게 진보적 의제를 관철시키지 못하는 것 아니냐" 정도의 시선이 자리하고 있을 겁니다.
이런 시선에 동의하는 것과는 별개로, 사실 잘 이야깃 거리가 되지 않는 것 중 하나는 "과연 한국에서 성적 소수자를 보호하자는 주장은 얼마나 정치적으로 인기가 있을까 (없을까)?" 입니다. 물론 대중 정당의 후보들이 이 주제와 일정한 거리를 두려 노력하는 걸로 봐서는, 그리고 우리 주변을 둘러봤을때는, 아마도 소수자들에 대중적 인식은 상당히 부정적일 것이라 누구나 추정하실겁니다. 다만 그 '상당히'가 도대체 얼마만큼 인가에 대한 이야기는 별로 없는 것 같습니다. 물론 게으른 제가 열심히 찾아보진 않았지만.
그래서 잉여력을 발휘해서 데이터를 뒤져봤습니다. 결론부터 말씀드리면, 한국 사회는 성적 소수자에 대해 '매우' 부정적입니다.
데이터는 World Value Survey (WVS)를 썼습니다. WVS가 완벽하지는 않지만 개인의 정치사회적 인식에 대한 학술 데이터로 워낙 광범위하게 쓰입니다. 또한 수시로하는 국내 여론조사에 비해서는 그나마 샘플링 등의 측면에서 문제가 덜 한 편입니다.
80년대 이래로 전세계에 걸쳐 거의 5-6년에 한 번씩 데이터가 쌓이고 있는데요, 지난 두 번의 wave, 그러니까 2000년대 중반, 그리고 2010년의 데이터만 썼습니다. 한국사회는 빠르게 변하고 있으니 아마도 더 오래된 자료는 오늘날의 현실을 보여주는데 크게 도움이 안되겠지요.
WVS 데이터는 누구나 다운 받을 수 있고, 아래에 제가 사용한 간단한 Stata 코드를 첨부했으니 관심있으신 분은 언제라도 스스로 분석해 볼 수 있습니다.
1. 비교적 관점에서 한국 사회의 성소수자에 대한 인식
'동성애를 정당화 할 수 있는가 (justifiable)?'라는 질문에 대한 변수를 썼습니다. 이 변수는 0 (절대 정당화될 수 없다) 와 10 (절대 정당화 될 수 있다) 사이의 스케일로 구성되어 있습니다. 나라 별로 평균을 냈을 때 한국은 전세계적 수준에서 사실 딱 중간에 있습니다.
글씨가 작아 잘 안보이는데, 그래프에서 화살표로 표시된 것이 한국 입니다. 위치만 중간이 아니고 실제 수치도 딱 세계 평균 (3.3)에 근접합니다. 아주 폐쇄적인 중동의 종교국가들도 모두 골고루 포함되어 있음을 감안하면 사실 상당히 낮은 수준이라고 할 수 있습니다.
2. 국내 수준
국가 수준의 평균으로 비교하는 것은 물론 한국정치에 직접적인 함의가 있다고 볼 수는 없습니다. 그래서 한국 샘플만 따로 들여다 봤습니다 (n=2392). 비록 평균이 낮긴 하지만 상당한 사람들이 성적 소수자를 긍정적으로 보고 다만 중간층이 적어서 평균치가 낮게 나왔을 가능성도 있기 때문이죠. 결과는 안타깝게도 그렇지 않았습니다.
앞에서 말씀드렸듯이 낮은 점수는 동성애에 부정적, 높은 점수는 보다 포용적인 의견을 의미합니다. 그래프에서 간명하게 볼 수 있듯이, 한국에는 동성애를 아주 부정적으로 보는 사람들이 절대 다수 입니다. 0-5에 해당되는 사람들을 '부정적'으로 분류 할 수 있다면, 샘플의 86%가 동성애에 부정적입니다. '심하게' 부정적인 사람들, 즉 0 혹은 1이라도 대답한 사람들만 쳐도 전체의 50% 정도가 됩니다.
반면 동성애에 보다 포용적인 사람들은 사실 극소수입니다. '적극적'으로 포용적인 사람들, 즉 9 혹은 10이라고 대답한 사람들은 전체의 3% 밖에 되지 않습니다.
정당 지지 별로 보면 좀 다를까요? 조사 시기에 리버럴이나 진보라고 할 만한 정당들은 민주당, 열린우리당, 민주노동당 정도였습니다. 이들 정당에 투표하겠다고 대답한 사람들과 다른 정당들 특히 한나라당이나 자유선진당 같은 보수당에 투표하겠다고 한 사람들 사이에 차이가 있었을까요? 답은 그렇다와 그렇지 않다 둘 다 인듯 합니다.
이 그래프는 정당 지지성향 별로 위의 동성애에 대한 의견 평균치를 나타낸 것입니다. 막대 그래프 위의 작은 뿔 같은 것은 95% 신뢰구간을 의미합니다. 확실히 진보-리버럴 정당에 투표하겠다는 사람들은 그렇지 않은 사람들보다 동성애에 더 포용적인 시각을 가진 것 같습니다. 95% 신뢰구간이 겹치지 않기 때문에 그 차이도 통계적으로 유의미합니다 (t-test statistics = 6.21).
그러나 이 차이는 전체적으로 정말 미미해 보입니다. 두 그룹의 차이는 1 (2.9:3.9) 정도이니까요. 다시 말해서 비록 진보-리버럴 정당에 투표하겠다는 사람들도 전체적으로는 동성애에 상당히 부정적인 시각을 가지고 있습니다.
데이터를 들여다 보며 든 생각은 문재인 후보가 이 부분에 대해서 확실하게 진보의제를 관철시키지 못한 것이 어느 정도는 이해가 간다는 것이었습니다. 진보-보수를 막론하고, 한국 사회의 동성애에 대한 시각은 극단적으로 부정적입니다. 단순히 특정 종교나 사회 세력의 사보타지가 아니더라도, 이렇게 전면적으로 부정적인 이슈는 선거를 치르는 누구라도 피해가고 싶지 않을까 생각합니다.
물론 용감하게 진보의제를 관철시킨 심상정 후보는 이런 측면에서 칭찬받아 마땅합니다. 그러나 아직도 치열하게 표싸움을 하고 있는 대중 정당의 후보가 이렇게 위험도 높은 선택을 하지 않았다는 건 비록 진보적인 가치에 동의하는 사람들이라고 해도 어느 정도는 이해해줘야 하지 않을까요?
아래는 데이터 보는 데 사용한 Stata 코드 입니다.
---------------
* use the integrated WVS data
gen gay = F118
replace gay=. if F118<1
gr hbar gay if S002>4, ///
over(S003, label(labs(vsmall)) sort(gay) ) scheme(s1color)
keep if S003==410
hist gay if S002>4, den scheme(s1color) xtitle("") ///
kden saving(kor_gay_dis, replace)
gr export kor_gay_dis.png, replace
gen temp = E179
replace temp=. if E179<0
gen liberal = 0
recode liberal 0=1 if temp==410002
recode liberal 0=1 if temp==410004
recode liberal 0=1 if temp==410005
recode liberal 0=1 if temp==410006
replace liberal=. if E179<0
ttest gay if S002>4, by(liberal)
cibar gay if S002>4, ///
over1(liberal) ///
graphopt(scheme(s1color) ///
legend(order(1 "conservative" 2 "progressive-liberal")) ///
ytitle("") ylabel(0 5 10) saving(ideo, replace))
gr export ideo.png, replace
아래는 데이터 보는 데 사용한 Stata 코드 입니다.
---------------
* use the integrated WVS data
gen gay = F118
replace gay=. if F118<1
gr hbar gay if S002>4, ///
over(S003, label(labs(vsmall)) sort(gay) ) scheme(s1color)
keep if S003==410
hist gay if S002>4, den scheme(s1color) xtitle("") ///
kden saving(kor_gay_dis, replace)
gr export kor_gay_dis.png, replace
gen temp = E179
replace temp=. if E179<0
gen liberal = 0
recode liberal 0=1 if temp==410002
recode liberal 0=1 if temp==410004
recode liberal 0=1 if temp==410005
recode liberal 0=1 if temp==410006
replace liberal=. if E179<0
ttest gay if S002>4, by(liberal)
cibar gay if S002>4, ///
over1(liberal) ///
graphopt(scheme(s1color) ///
legend(order(1 "conservative" 2 "progressive-liberal")) ///
ytitle("") ylabel(0 5 10) saving(ideo, replace))
gr export ideo.png, replace
Saturday, November 26, 2016
Antiauthoritarianism in the US and South Korea
A paper topic I've been thinking about in the past several years is Anti-authoritarianism around the world and its decline (that is, the rise of authoritarianism). As always, procrastination got in my way and I really haven't done anything on it. Recently, perhaps because of the election, it seems like an outcry of the rising authoritarianism in developed societies is not very uncommon. While I thought the argument has its merit, I wasn't really thinking of looking at the data even though I had a variable (the name is, you know, "anti-authoritarianism"), part of which was a central dependent variable of one of my published articles.
In my attempt to overcome Thanksgiving food coma this afternoon, I plotted a cross-national bar graph for this variable.
The variable antiauth captures individuals' Anti-authoritarianism (opposition to a military- or strongman rule) and each bar represents national average of the variable in each 'wave' of the Survey.
Both United States (yellow) and South Korea (green) showed their strongest anti-authoritarianism in the 1990s, which subsequently declined significantly in the 2000s. Recent presidential election results in both countries should make a certain amount of sense. To be fair, I didn't run numbers in any systematic way to see if these differences are actually statistically significant (yet).
I should really start working on this (after submitting what I've been working on, that is).
In my attempt to overcome Thanksgiving food coma this afternoon, I plotted a cross-national bar graph for this variable.
The variable antiauth captures individuals' Anti-authoritarianism (opposition to a military- or strongman rule) and each bar represents national average of the variable in each 'wave' of the Survey.
Both United States (yellow) and South Korea (green) showed their strongest anti-authoritarianism in the 1990s, which subsequently declined significantly in the 2000s. Recent presidential election results in both countries should make a certain amount of sense. To be fair, I didn't run numbers in any systematic way to see if these differences are actually statistically significant (yet).
I should really start working on this (after submitting what I've been working on, that is).
Wednesday, November 16, 2016
Presentation Template
So we're getting closer to the end of the semester, which means there should be a lot of student presentations going on. In my GLOA 605, which is an MA-level interdisciplinary research method class, this is a big deal. For two weeks, students present their research projects. Having taught this class before, I noticed (and found it somewhat inconvenient) 1) that they have very different styles of presentation which sometimes undermines fair grading and 2) that many students had never presented their work in a professional manner (!) and thus messed up with time management due to poor structure of the slides. That led me to make a template that everybody in class has to adhere to. It might sacrifice their creativity a little bit, but it's a formal training through which everybody is expected to garner certain types of skills.
Monday, September 26, 2016
Why does the recovery in the developed world not translate into the growth of the developing world?
Short Answer: the recovery in the developed does not mean the recovery of their demand for imported goods. A, rather old, report from the UNCTAD provides us with a very intuitive graph. The positive associations between the growth of the North and the growth of exports of the South (which is considered a major conduit for their desperately needed growth) had disappeared in the 2012-3 when the recession took finally took its toll.
Analysts should (and must) have known this coming because this shift had a long way to come; and yet, not much of preparedness have we witnessed. Indeed, the disastrous Hanjin bankruptcy would have been negated, if not avoided, if this sort of intuition was reflected in policymaking.
Friday, April 22, 2016
What I am going to talk about Monday.
Until very recently, I had been declining the request to give a talk about Asia for various reasons. Sometimes the schedule simply didn't work out. Sometimes I had absolutely no idea about the topic. But then I felt pretty bad about refusing to go all this time and finally accepted participate in a student-organized project.
I still don't know what I will be talking about. but this is the slides (will be revised a lot, probably). I really don't think there has been any `pivot'. But the underlying conditions suggest that Asian countries will have to make some choices pretty soon. And the data suggests that keeping status quo (not really using more RMB) would be the right choice.
I still don't know what I will be talking about. but this is the slides (will be revised a lot, probably). I really don't think there has been any `pivot'. But the underlying conditions suggest that Asian countries will have to make some choices pretty soon. And the data suggests that keeping status quo (not really using more RMB) would be the right choice.
Wednesday, April 13, 2016
[book review] Demanding Devaluation by David Steinberg
In Demanding Revolution, David Steinberg proposes a convincing theoretical framework, namely,``conditional preference theory'', to answer the question, `why undervaluation is so rare despite its unequivocal economic benefit?'. He suggests that strong manufacturing sectors support undervaluation when and only when the cost the policy incurs to them is reduced by the government. These costs, including increasing wage levels and imported input costs as well as decreasing access to credits, can be reduced when 1) financial institutions are controlled by governments (such that banks are forced to provide cheap credits -- ``financial repression'') and/or 2) labor rights are limited (such that upward wage pressures are suppressed).
The empirical evidence he presents lends strong support to this framework. Using a mixture of quantitative and qualitative evidence, he shows that undervaluation is most likely in countries where 1) manufacturing sector accounts for large parts of the national economy, 2) labor rights are limited, and 3) banks are, at least partially, nationalized. While the quantitative evidence neatly illustrates this evidence in a cross-sectional time-series manner (although I believe Figure 2.3 can be drawn better), the beauty of the empirical analysis of the book is in the case studies. Of the five country cases studied, China and Korea confirms one side of the story, namely that sustained (though intermittently interrupted) periods of undervaluation was made possible by strictly controlled labor and banks combined with large manufacturing sectors. The cases of Argentina and Mexico, on the other hand, illustrate the prevalence of overvaluation when the condition present in China and Korea are absent (and the policy debates were overwhelmed by extraordinary levels of inflations) even though manufacturing sectors were fairly strong. The Iranian case represents a rather odd scenario where manufacturing sector is weak and commodity export is dominant in the national economy.
While I thoroughly enjoyed the book, a few aspects of it struck me as unclear, asking for further investigations. First, what was the role of the state, particularly in the cases of China and Korea? The conditional preference theory essentially is interest group theory, assuming that the strong interest groups' preference translates into policies without much interference of the state. But the very conditions that enabled undervaluation in these cases--nationalized banks and controlled labor--is the products of strong, autonomous states. Indeed manufacturing sectors were nurtured by states in its infant stages and elevated to the `strong' status through deliberate government actions; undervaluation was one of those. Granted, in the later stages, that is the 1990s for Korea and 2000s for China, industrialists outgrew the influence of the state and rather dictated exchange rate policies in some ways. This was not true, however, for the majority of the times that the book covers. (To be fair, the author does address this issue here and there, but I think this state-society(business) issue needs to be more seriously handled, preferably in the theory section)
Second, in a similar vein, the external factors are not modeled in the conditional preference theory. The big difference between China/Korea and Argentina/Mexico, however, is their industrialists' access to foreign credits. In the former, individual firms' foreign borrowing had been substantially limited until financial liberalization (which really hasn't come to China even these days) and the degree to which the global interests rates influenced domestic markets was much narrower in these countries than in Argentina and Mexico. Again, the book does explain this part in the empirical section, but more as a sideway show that occasionally conditions each state's policy choices, not as a main independent variable.
Overall, the book significantly extends our understanding of monetary policies of developing countries. The literature does not really go beyond ``competent Asian states = undervaluation / incompetent Latin American-African states = overvaluation''. The book provides much more in-depth theorizing with convincing empirical evidence. I would use this for my graduate seminar.
Subscribe to:
Posts (Atom)





