Autocratic Regime Data

In light of my previous post, I found this Autocratic Dataset from Barbara Geddes and Joseph Wright. The data was amassed meticulously as they constructed the different nature of autocratic regimes. This new measurement of autocracy, which sometimes used interchangeably with authoritarian, could advance future research avenue on autocratic durability and unfold many other interesting questions. For comparison, you may want to look at the binary regime data and Przeworski et al political and economic dataset.

Did Stable Regimes Affect Development?


I was supposed to work on my paper today. But I randomly came across the data from world development indicators of World Bank which I found very useful. There are thousands of variables which come in handy. You can do interactive comparison on selected variables and countries of interest (you can also adjust the chosen years). So I compared the growth of Indonesian GDP per capita (at nominal level) from 2000 – 2014 to other “commensurate” countries in SE Asia. I added to all developing countries in East Asia and Pacific to get a sense of the rates of overall growth of the region. Developed countries such as Japan, South Korea, Singapore are excluded.

One thing that pops up directly is that how the growth of per capita GDP in East Asia and Pacific outweighed all countries in SE Asia throughout the years. The gap was even wider before the coming of the financial crisis in 2008. Then the crisis kicked in and all countries suffered so badly. The Philippines,  Malaysia, and Thailand were the most affected countries while Indonesia and Vietnam were relatively ok. But interestingly, a year later these three countries rebounded enormously and overpassed Indonesia and Vietnam in 2010 (yes, both Thailand and the Philippines fell down again greatly in 2011). All countries in the region had been recover in 2010, but then the rates of growth were in steady declined. And since then the rates tended to converge at the same pace.

The paper that I am supposed to write seeks to make sense the effects of regime types, if any, on income distribution. But this graph puzzles me on another thing, which doesn’t necessarily address to the question in my paper: To what extent we can theorize the relationship between stable regimes and certain economic development indicators (such as GDP growth, income per capita, income distribution, etc?). Of course before we proceed, I need to clarify what do I mean by “stable regimes.” I cannot think of any data or articles that address this term specifically. So I’d rather think about it on my own.

I intuitively think about stable regimes in terms of the ways the regimes exit or not exit power at certain periods of time. They could either be democratic or authoritarian countries. For authoritarians, it could be soft authoritarian like that of Malaysia or Singapore and probably Vietnam, which always been governed by the same regimes, or Indonesia under Suharto’s rule. For democrats-alike countries, we have to look at the ways the transfer of power have been done, was it really peaceful or not? Some democratic countries underwent democratic drawbacks and some other experienced reversal to authoritarian government. So it’s important to note the baseline years of classifying “stable democracies”; how many years countries could be counted as stable are more, I think, a matter of choice (one can say 10 or 15 years, or based on a minimum number of peaceful elections that have been held). On the other hand, the “unstable regimes” would be countries in which the transfer of power ensued either forcibly (like via coup or war, for example) or through regimes overthrown (be it to democratic or authoritarian reversal). Again, the definition of stable and unstable regimes will be hinged on the extent of time we’d prefer to employ. One country can be classified as a stable regime, but also can be counted as an unstable in any other time.

Once the data holds, then we can test the relationship of this stable regimes to our economic indicators of interest. Many big questions in comparative politics can be drawn from this. For example, what are the differences of growth in stable authoritarians and stable democracies? To what extent regime types affect the levels of development? Or what conclusions can be raised about country’s levels of development when one country has an authoritarian system in the past compared to its current democratic system (like Indonesia?).

An Issue in American Politics

At times, I wonder how distinct actually the field of American politics (AP) to comparative politics (CP). Yes, the focus of analysis of the two is obviously different. AP, the name speaks for itself, focuses on political dynamics in the United States. One seeks to study AP, for example, will be socialized with the ways American democracy works. They can study congress or presidency, how gender and race affect political discourses, public opinion, or voting behavior. On the other hand, the focus of CP potentially intersects with the broad themes in AP. None of the subject of interests in AP are alien for CP. Instead, the topics of CP are quite broader (i.e very few political scientists have interest to study social movement or leftist ideology in the US, whereas these topics have been scrutinized in great details by comparativists). Having taken both CP and AP, I can tell that the considerable area of interests between the two are not that different.

But the distinction between the two subfields is more obvious if we compare the requirements for methods courses. To get you a sense, here are brief descriptions for AP and CP at the Ohio State, one of top-notch PhD programs in the US. As you can tell, there are minimum FIVE courses on methodology (from quantitative methods to formal model) that must be taken by those who study AP. Whereas in CP, the requirements are somewhat modest. Students are usually expected to take the combination of research design and quantitative class, aside from a foreign language requirement of the country of interest (which is, predominantly, has been eliminated in many schools).

So as you might expect, in the first place, I can say that the field of AP is mostly method-driven than CP, which is largely, I suppose, a problem-driven field. AP is methodologically more rigor and more quantitatively oriented field than CP (Ian Shapiro has an excellent discussion on this). This is not to say that such a methodologically oriented field is wrong. That’s not the point. But I was just wondering how isolated AP could be if scholars engage in AP debates don’t put their findings in comparative perspective. This idea also calls into question how compliment or how possible the scholars in the two subfield can communicate each other, given boundary in their respective fields (more discussion on this Linz & Stepan and Przeworski).

Working Paper

I just posted my working paper to SSRN. This paper has been presented once in research colloquium at OU. Though I halfheartedly felt actually if a kind of research network like SSRN can be beneficial for a stuff I am working on, I’ll be happy to hear any comments. I wrote this paper for an independent study last Fall. Here is the abstract


This paper explores the bases of people’s democratic preference in mainly nondemocratic countries of Southeast Asia. It does so by, first, classifying seven Southeast Asian countries according to their governmental systems. Drawing from the works of democratic theorists and leading democracy indices, the paper explores four different forms of government in the region, namely, democracy, competitive authoritarian, the drawback or unstable democracy, and the communist system. Second, given their different forms of government, the paper then proceeds to seek possible explanations on people’s democratic support. Using the dataset from Asian Barometer Survey Wave 3 (2010-2012), there are four hypotheses to be tested, which are, economic explanations, social capital, modernization theory, and the degree of ethno-linguistic and religious fractionalization. Employing multinomial logit regression, the findings suggest that while democratic support is evident in seven countries across the board (and the support tends to be higher in nondemocratic rather than in democratic countries themselves), the democratic support in nondemocracies has largely been restrained by (1) their conditional economic factors, (2) the low level of trust among people in general, and (3) the ambivalent stances of the middle class to align with democratic principles