Many readings for this week concern the functions of institutions under dictatorship, and most of them argue that they contribute to authoritarian survival. To be honest, I really like this idea though it is true institutions do not always prevent authoritarian regimes from collapse but foster democratic transition (even it is the case, Babara Geddes will say autocracies with democratic institutions are more likely have peaceful transition process) as Schedler aptly indicated.
One reason for my fondness is for a long time many scholars tend to treat institutions under autocracy as the facade of dictatorial rule thus devoid of academic significance. Take elections as an example. While many American researchers have been wondering and questioning why elections are held in China even though people are not very interested in them as they know those backed by the party will win anyway under most circumstances, many political scientists working on democratization may disdain the necessity of a systematic study of Chinese (rural) elections or even presume they will pave the path of democratization in China. Therefore, I really like the way Magaloni sets up her book, by which she justifies how elections under single-party autocracies can help the PRI coordinate internal competition and coopted opposition that in turn sustained its rule. Similarly, in Gehlbach et al., Gandhi et al;,and Smith's work, they all more or less claim that (quasi-)democratic institutions act as a critical platform for autocrats to survive or promote economic development. Their ideas encourage my research on China's ethnic autonomous system that I am currently working on. Instead of treating the party's provision of autonomy and policy privileges as the result of Lenin and Mao's conceptual compromise between their beliefs in Marxism and the western principle of national determination, a decentralized autonomous system to certain degree helps both regimes (China and the USSR) incorporated different ethnic groups into a unified socialist nation-state/empire.
However, based on my study of the communist ethnic autonomous system, I do think Gandhi and Przeworski's model can be refined in several ways. First of all, they seem to treat the organizational capacity of the opposition as a variable that will determine autocrat's willingness to make concession (a summary is in p.2), what if we treat the capacity of autocrats as unstable in the meantime? In other words, the changing capacity of autocrats to repress rebellion may also play a role in the process of making concession. The capacity of the ruling group under autocracies may be seriously compromised when the ruling group (i.e. the Chinese Communist Party) suffers from serious internal power struggle (especially after the death of a important leader, e.g. Mao). Therefore, I will say autocrats tend to make more concession when they know they are not strong enough to repress the rebellion. This argument may also explain why given the existence of a credible threat of rebellion (e.g. the presence of ethnic secessionism) the party autocrats across time reacted in different ways. Also, a minor point for them is using the number of party may not really capture the degree of concession autocrats have to make as concession may be significant when there is a large party composed to many competing factions.
For Smith, I really like his piece as he tackles the variation in regime durability of the single-party autocracies. His theory fits many cases very well, and his selection of cases is also well explained. The presence of an organized opposition and the lack of access to rents do more or less provide the incentives for autocrats to establish a well-organized party, and this argument fits the case of Taiwan. However, I would like to push his argument further in several ways. One, where does the opposition come from? Can it be a foreign threat? (e.g. Taiwan against Mainland China; Singapore against Malaysia (?) - I need Man Yan to fill me in here). Second, while Smith delineated the incentives for autocrats to build a strong party very clearly, I am wondering if he can talk more about the "means" of achieving so. He provides two cases, but there is not very systematic explanation about how autocrats manage to coopt opposition and broaden the pool of control across different social segments. Thirdly, how long does it take for a single-party authoritarian regime to be counted as "durable?"
My last thought regards Gehlbach and Keefer on investment without democracy. While I find their argument interesting, to be frank I am more disappointed after finishing reading this. Most important of all, what is "institutionalization of the ruling party?" I cannot really see a very good definition in their article. Do they mean the institutionalization of cadre recruitment and promotion or the institutionalization of the investing process (or both)? Even if it is the former as they implicitly suggest, I am confused as I do not really see how this resolves the collective action problem and promotes foreign investment in China. After all, many China scholars have long been concerned about the limited impact of bureaucracy institutionalization in China, and many of them have argued that it is not the changes in the cadre system but the changes in the incentive mechanism that encourage local cadres to seek economic growth and foreign investment for their own sake (that is exactly Oi's argument, which is also shared by B. Naughton). In this sense, stabilizing a cadre's career path has much less to do with economic development (instead, people focus more on how the CCP has been recruiting people with professions, such as finance, engineering, law, business management, which are relevant to economic development). So I guess the question will be: How will Gehlbach and Keefer's argument relate to existing studies of China's economic miracle?
Though much of what I want to say is in my referee report on Jessica Week’s paper, I want to reiterate here how much I enjoyed this article. I appreciate her attempt to argue that it is not just democratic leaders that face domestic punishment after foreign policy blunders such as losing a war. While I think there are various minor theoretical and methodological issues that should be resolved in further drafts, her work on identifying variation within autocracies that make some states more or less vulnerable to domestic pressure in foreign policy decision-making is interesting and fills an important gap in the international relations literature.
I also enjoyed Magaloni’s book, which was an interesting example of how elections, rather than indicating democratic consolidation, can serve to strengthen semi-authoritarian and authoritarian regimes already in place. This is certainly true in post-Soviet states, and Levitsky & Way (2002) have written some useful pieces that use façade elections to help them classify variation within authoritarian regimes. I think there are other mechanisms at work in the use of elections as a way to consolidate authoritarian rule that Magaloni doesn’t identify such as the fact that the elections can serve as an assertion of power of the populace, as the tactics used to render elections not fair and free can also emphasize the weakness and disenfranchisement of both the general population and the opposition. In demonstrating the weakness of the opposition, the regime underscores its position of power in society and legitimates itself both internationally and domestically. Further, by recognizing the international norm of elections as a pathway to democracy, the regime burnishes its own credentials and can use the rhetoric of democracy to achieve international standing and recognition as a “transitioning” state. More subtly, elections serve to strengthen semi-authoritarian regimes through the gauging of political support in a one and half party system and by strengthening patronage networks in the run-up to the elections Additionally, leaders can use the election has a means to institute un-free laws and institutions such as media and freedom of assembly laws. Even more nefarious techniques such as voter intimidation and increased police presence in opposition can also be used to institute an atmosphere of fear and trepidation in the run-up to an election. In contrast, leaders might also use the election time to appeal to the populace by starting long-overdue infrastructure projects or by handing out money and patronage. Another important aspect of elections in un-free states is that fact elections of this type are rarely issue based and instead become highly personalized and lacking in content. This type of electoral system removes politics from political issues and transforms the election into a personal advertisement for the autocrat and his regime. In these regimes, political parties are often centered around a small elite group of actors or a charismatic leader, and lack coherent ideologies. Since ideology has been removed from the election, the only constant in political life is the attitude toward the government, which is then confirmed by landslide election results in favor of the incumbent.
I find it interesting to compare and relationate the explanations given for the Mexican PRI's longetivity by both Haber and Magaloni (although I am definitely not a Mexicanist expert!). Whereas Haber emphasizes the importance of the logic of organizational proliferation, Magaloni's model (built on the Mexican case) gives a threefold arguments: behavioral incentives for elite unity, the strategic support of voters and the opposition's coordination dilemmas.
The explanation of Haber is not extensive, but I believe it at least partly meets with the first and third dimensions of Magaloni's model. However, it seems to me that it does not recognize any importance for the electoral or even broader mass support to the autocracy. Neither does so the logic of terror described by Haber.
Personnally, I've been convinced by the importance of electoral/ mass support's argument of Magaloni, since the regime can't repress "so massively". One could qualify this support as falsified (as Kuran 1991, referred to by Magaloni, does it for Communist regimes); but I believe this does not diminish its importance as an explanation factor for the survival of autocracies/ hegemonic-party regimes.
I also enjoyed Magaloni’s work. While it was echoing a lot of Geddes, it answered a lot questions that arouse when reading the former, such as the ways in which elites and others are co-opted into the ruling party. Of her independent variables that explain how one-pary regimes survive and transition our of power—electoral support, electoral fraud, elite fragmentation and coordination challenges—I kept thinking about mass electoral support. One party regimes need a large turnout of voters and a large margin of victory at the polls to both deter challenges to their rule and the need to include any opposition groups that do exist. I started thinking about this but on a none-marco level. In the last legislative election in Lebanon, Hezbollah had the largest get out the vote drive, which included vote buying and intimidation, of all the parties even though their districts were largely uncontested for precisely the same reasons: they wanted to show that they represented the Shia voice in Lebanon, obviate any potential rivals from the traditional Shia families who resent for diluting their political power; and discount the need to incorporate other disparate views that lie within their districts, especially those of the communist party. At the same time, the party has established a good rapport and relationship with the other key Shia party, Amal, to the extent that they are considered to be co-opted. All these traits lend it to being a hegemonic one party actor among the Shia population of Lebanon. I also found the explanation to of why these regimes transition to democracy interesting and convincing.
- First, I think it demonstrates the role for structured comparative case studies as a way of thinking about causality in small-N, comparative politics that is poorly captured by regressions (although the paper starts off with an all but meaningless replication of a 17-variable regime regression).
- Second, it allows for choices being made on different stages in a development: There are the initial conditions (struggle and rent), but there is also the question of what to do with rent if it appears later on, or how to deal with opposition later on.
However, there are also two things I don’t like:
- First, I don't think he elaborates enough on the possible theoretical strengths of this approach, and he is too eager to present clear narratives that fit the simplest variant of his theory - so simple, in fact, that it could be reduced to a regression (dependent variables would be long and hard civil war/struggle at the time of party strategy decision and access to rent). This seems unnecessary, since his cases presents a more complex picture: For instance, he claims that in Indonesia, the regime decided to use oil windfalls to deepen the party, rather than buying off people.
In terms of thinking about causality, such a two-step model, tracking mechanisms between initial conditions and outcome, is interesting. One could assume either that initial conditions, e.g. a deeper party structure increases the likelihood of subsequent rent to be used to strengthen the party (which Smith seems to do), or one could treat the relationship as random. Thinking of causality in terms of numerous steps connected through specified or unspecified probabilities might often more closely resemble how we think about causality, but often renders regression analysis uninterpretable. I think Smith’s narratives point to this, but it is not explicated.
- Second, the paper also shows some of the dangers of doing case-studies to confirm a theory: There seems to be a clear confirmation bias, i.e. in interpreting events in line with the theory. For instance, he seeks to portray PAIGC as a weak party, and seriously downplay the role of Portuguese forces in the country. First of all, although it is true that fighting was harder in other Portuguese territories, the opposition PAIGC met was arguably significantly stronger than what Nyerere encountered with the more peaceful British in Tanganyika. Second, Portuguese efforts increased over time, and in the late 1960s and early 1970s, they were pretty intense.
The bad choices made by Luiz Cabral are interpreted as excluding groups, whereas Nyerere’s are interpreted as including them – although there is probably some truth to this, it is striking that Nyerere’s dismissal of the existing elite (who were suspected of not being nationalistic) is not seen as exclusive, whereas groups excluded by Cabral are.
Likewise, Marco’s access to forestry revenues is seen as a huge boon, whereas Indonesian oil revenues prior to 1973 are just mentioned in passing as smaller than in 1960. By simplifying in this manner, I think Smith unnecessarily downplays the role of choices made by leaders at least partly independent from initial conditions.
I thought Kalyvas's article on the breakdown and decay of Soviet regimes was interesting, though perhaps his argument would be different today. While his point about the differences between breakdown and decay are well taken, the theory overall seems a bit dated, as today the variance between the former Soviet republics grows. Additionally, his comments on political competition no longer seem valid, as many of those who had been voted from power are now back in power, and if not former apparatchiks, at least their protege from the Komosomol. The description of decay and breakdown are thought-provoking, but do not they fall short because they do not entirely explain the reemergence of authoritarianism in much of the former Soviet Union, nor the diversity of outcomes. The patterns of decay and breakdown, while similar, must have some obvious differences that should be further detailed in order to make sense of current former post-Communist regimes.
I rather liked Weeks' paper, and I have to admit I was deeply envious of her army of research assistants who did all this wonderful archival work and coding. Her exposition was convincing, she frankly admitted the limitations of what she was doing, and she provided work arounds for possible problems in the coding.
Nonetheless, I wonder if she couldn't have pushed her "war" variable a little more. Trimming the COW data to include only large or cross-border conflicts is all well and good, but perhaps there ought to be an additional distinction between limited and total wars, which could presumably be coded the same way as regime type (ie by using newspaper reports and academic accounts). Because it seems intuitively plausible that there are relevant differences between a minor border skirmish that didn't work out and losing World War II. Good meaures might be lost territory or some significant and costly concession.
I limit myself to 3 articles because I realized my comments were becoming a bit long...
a) Jessica L. Weeks
Here I have to be more critical than Robert. A critical point she doesn’t mention –as far as I remember- is that conventional wisdom is in fact true when we compare the ousting of autocratic leaders to democratic. It only doesn’t hold for “constrained” authoritarians. If I got this wrong, please correct me. If it is right, the paper’s contribution remains but is limited to a finer grain approach to autocracy, not to contradicting conventional wisdom.
An easy criticism –but not for that reason less valid, is that cross-tabulations are even more limited than regressions in determining causality. Therefore, in the paper we are given “empirical patterns” due to the small-n common problem. (Being a bit distrustful, could it be that regression results were not actually supporting her thesis?)
Finally, I have to agree with the author’s concern on page 35: “endogeneity between the leaders’ fates and the regime type codings is a real possibility”. In summary, this paper does not pretend to establish causality using methods like instrumental variables and instead builds a plausible argument with tentative results.
b) Benjamin Smith
It is a solid study that, as the author admits, needs to be tested. It reminded me of another comparative study by Clifford Geertz (Islam Observed, Religious Development in Morocco and Indonesia). I believe both authors make a good use of Mill’s method of agreement –that is, they compare disparate cases which have a commonality that is presumed the cause of the phenomenon. In Smith’s case, the cause would be the original strength of the party.
My critique is related to the graph on page 432: could something be more path-dependent? I understand/hope it is a simplification, and yet I can compare to game-theoretic assumptions that are just as little realistic.
c) Stephen Haber
I might be biased here because I very much enjoy the work of B. Weingast, not the least Violence and Social Orders. He uses technical concepts in a language understandable to undergraduates. I do think he reconciles the economics and sociology tradition simply by noting that the 3 winning strategies –which are 3 out of the 4 equilibria of the game- can be turned into categories. The 4th equilibrium is the dictator being ousted –certainly not a winning strategy from the dictator’s point of view…
I thought it was funny enough that Acemoglu et al. correct the econometric “misspecifications” of Przeworski once again, especially after reading Epstein’s et al. article in AJPS two weeks ago (see page 9).
A (minor?) critique to the paper (page 10) is whether the discontinuity factor (delta) grows with the dictator’s age. Is everyone like Mugabe? Or are some dictators less predatory as they grow older? That relates to a larger concern of the paper being very focused on the “predatory nature” of dictators, maybe ignoring that often times formal and informal constrains don’t allow them to be –like Tunisia nowadays.
I was supposed to email everyone but I couldn't find a way of sending an email to the class on courseworks...anyway my referee report on the Weeks paper is up on courseworks in the shared files folder. @ joan in regards to your criticism of Weeks not using regression analysis: regressions are not always the most appropriate methodological choice in political science! I think she is right to use a cross-tab in this case, using a regression wouldn't capture the correlation she was looking for. I think there is a real bias in terms of quantitative analysis and in some cases being more explicit with the logic and less reliant on the numbers can be a good thing in my opinion. Second, Weeks does challenge conventional wisdom: the conventional wisdom is that it is the very nature of authoritarian regimes (lack of accountability to a domestic audience) that makes them immune to audience costs and less vulnerable to ouster. Here Weeks shows that this is not the case, some authoritarian regimes are vulnerable to ouster because they are more highly constrained by the selectorate. Instead of treating authoritarian regimes as a residual category, she builds on the large spate of recent literature that distinguishes between these regimes. Further, rather than just creating a typology of variation within authoritarian regimes she actually uses this difference to show a pattern of policy vulnerability. In any case, this is all just my opinion and I may be biased by the fact that I think she is doing great work on a really fascinating area of research (my own research area as it happens haha).
@ Nils: Based on your comments, I would like to say something more about Smith. First of all, I am wondering if he can be more specific about the so-called "co-optation" through the establishment of a dominant party? Will there be any difference if the party is meant to co-opt the elite (as opposed to autocrats) or the mass? It will be surely good if Smith can focus something more on the process of party-building in these authoritarian regimes. Second, I think it is really hard to justify the selection of cases even though Smith does explain a little bit before he starts his case studies. How should we do this? I am also inclined to question if his selection is really appropriate. Thirdly, is Smith's article on the formation of dominant parties or the durability of dominant parties? This point is rather minor but may be worth discussing.
First I would like to say that from the literature related to the survival of the PRI I find that Magaloni builds a solid case for showing that elections were not devoid of content and its only function was to legitimize the regime. On the contrary elections were a very efficient information collection mechanism to be able to selectively coopt elites, union members and even the opposition.
In general I agree with the argument she presents however I would like to add some facts that I believe may help us understand the survival of the PRI better
1) While it is true that the populist wing and the neoliberal wing had different approaches as to how to handle the economy, it is also true that when it came to co-optation and clientelistic practices, both wings operated in the same way. For example, when Salinas (1988-1994) made explicit his intentions to reform, some union leaders were not willing to cooperate. Salinas' response was to incarcerate the leader of the oil workers union aka "la quina" and to change the leader of another important union the teachers. Therefore, signaling that he was the president and ultimately his decisions were law.
2) I don't think Magaloni gives sufficient attention to the effects that signing NAFTA had on the Mexican system. Signing NAFTA required demonstrating that the country could at least have stable macroeconomic indicators thus, by the time Salinas was trying to enter into this treaty, he had significant pressure from abroad, thus restraining to some extent the policies he could explicitly carry.
3) I am not sure if a regression is necessary to demonstrate that PRONASOL was in fact an efficient instrument for vote buying. I believe that the evidence presented in a case study is sufficient for proving that PRONASOL was designed to buy back defectors to the PRD. For instance, I would add the fact that Salinas' presidential candidate was the Minister in charge of administering this program. In addition, presidents had "reserved funds" that they also used to keep their co optation system.
In line with Magaloni I also provide some phrases from the Mexican politicians that I believe have great explanatory power:
"I don't see them, and I don't hear them" Carlos Salinas in reference to the opposition claiming fraud in the 1988 elections.
"I will defend the peso like a dog" Jose Lopez Portillo-the next day the peso was devaluated…
"If you think it don't say it, if you say it don't sign it, and if you sign it back off" Fidel Velazquez Union Leader
@ Joan: I think your points on Week's article are very interesting. Even though I am not an expert on regression analysis, I think it will be a nice try if you can do this. I do not think there is much to say which one is good before we see the results, but there is still a chance to tell which one may better support her argument. Also, yeah, I think her point in this article is not to contrast autocracy and democracy but to dissect the concept "autocracy" and to explore the variation within this category.
By the way, I totally agree with your comment on Schedler's article! While it is very straightforward and clear, I really think it can be arranged in a more readable way.
I really liked this week's readings, but something struck me as I was doing the readings: there is hardly any mention of the possible role of the bureaucracy in supporting authoritarian regimes in this week's readings. Although Smith has provided a cogent argument and strong evidence through 4 case studies that the origins rather than the existence of single-party rule are crucial in determining the viability of such regimes, there is no mention of roles that bureaucratic agents might have played in reinforcing single-party rule. However, the possibly significant role that bureaucracies might have played is hinted at by Haber in his paper, where he notes that "the dictator needs this launching organization, as well as the other organizations within the state that the launching group permeates, because without them he cannot run the country". Indeed, one could argue that any dictator or ruling party would necessarily depend upon bureaucratic agents with specific areas of expertise to aid them in their rule. Hence it might be interesting to consider how the relationship between the single-ruling party and the bureaucracy in authoritarian regimes could have possibly reinforced the viability of the ruling-party, e.g. whether the ruling party has awarded significant benefits to the bureaucrats to incentivise them to work for the party; whether bureaucratic drift could eventually lead the ruling party to a downfall because the ruling party alone lacks the necessary capabilities and expertise to run all aspects of the country.
Many readings for this week concern the functions of institutions under dictatorship, and most of them argue that they contribute to authoritarian survival. To be honest, I really like this idea though it is true institutions do not always prevent authoritarian regimes from collapse but foster democratic transition (even it is the case, Babara Geddes will say autocracies with democratic institutions are more likely have peaceful transition process) as Schedler aptly indicated.
ReplyDeleteOne reason for my fondness is for a long time many scholars tend to treat institutions under autocracy as the facade of dictatorial rule thus devoid of academic significance. Take elections as an example. While many American researchers have been wondering and questioning why elections are held in China even though people are not very interested in them as they know those backed by the party will win anyway under most circumstances, many political scientists working on democratization may disdain the necessity of a systematic study of Chinese (rural) elections or even presume they will pave the path of democratization in China. Therefore, I really like the way Magaloni sets up her book, by which she justifies how elections under single-party autocracies can help the PRI coordinate internal competition and coopted opposition that in turn sustained its rule. Similarly, in Gehlbach et al., Gandhi et al;,and Smith's work, they all more or less claim that (quasi-)democratic institutions act as a critical platform for autocrats to survive or promote economic development. Their ideas encourage my research on China's ethnic autonomous system that I am currently working on. Instead of treating the party's provision of autonomy and policy privileges as the result of Lenin and Mao's conceptual compromise between their beliefs in Marxism and the western principle of national determination, a decentralized autonomous system to certain degree helps both regimes (China and the USSR) incorporated different ethnic groups into a unified socialist nation-state/empire.
However, based on my study of the communist ethnic autonomous system, I do think Gandhi and Przeworski's model can be refined in several ways. First of all, they seem to treat the organizational capacity of the opposition as a variable that will determine autocrat's willingness to make concession (a summary is in p.2), what if we treat the capacity of autocrats as unstable in the meantime? In other words, the changing capacity of autocrats to repress rebellion may also play a role in the process of making concession. The capacity of the ruling group under autocracies may be seriously compromised when the ruling group (i.e. the Chinese Communist Party) suffers from serious internal power struggle (especially after the death of a important leader, e.g. Mao). Therefore, I will say autocrats tend to make more concession when they know they are not strong enough to repress the rebellion. This argument may also explain why given the existence of a credible threat of rebellion (e.g. the presence of ethnic secessionism) the party autocrats across time reacted in different ways. Also, a minor point for them is using the number of party may not really capture the degree of concession autocrats have to make as concession may be significant when there is a large party composed to many competing factions.
For Smith, I really like his piece as he tackles the variation in regime durability of the single-party autocracies. His theory fits many cases very well, and his selection of cases is also well explained. The presence of an organized opposition and the lack of access to rents do more or less provide the incentives for autocrats to establish a well-organized party, and this argument fits the case of Taiwan. However, I would like to push his argument further in several ways. One, where does the opposition come from? Can it be a foreign threat? (e.g. Taiwan against Mainland China; Singapore against Malaysia (?) - I need Man Yan to fill me in here). Second, while Smith delineated the incentives for autocrats to build a strong party very clearly, I am wondering if he can talk more about the "means" of achieving so. He provides two cases, but there is not very systematic explanation about how autocrats manage to coopt opposition and broaden the pool of control across different social segments. Thirdly, how long does it take for a single-party authoritarian regime to be counted as "durable?"
ReplyDeleteMy last thought regards Gehlbach and Keefer on investment without democracy. While I find their argument interesting, to be frank I am more disappointed after finishing reading this. Most important of all, what is "institutionalization of the ruling party?" I cannot really see a very good definition in their article. Do they mean the institutionalization of cadre recruitment and promotion or the institutionalization of the investing process (or both)? Even if it is the former as they implicitly suggest, I am confused as I do not really see how this resolves the collective action problem and promotes foreign investment in China. After all, many China scholars have long been concerned about the limited impact of bureaucracy institutionalization in China, and many of them have argued that it is not the changes in the cadre system but the changes in the incentive mechanism that encourage local cadres to seek economic growth and foreign investment for their own sake (that is exactly Oi's argument, which is also shared by B. Naughton). In this sense, stabilizing a cadre's career path has much less to do with economic development (instead, people focus more on how the CCP has been recruiting people with professions, such as finance, engineering, law, business management, which are relevant to economic development). So I guess the question will be: How will Gehlbach and Keefer's argument relate to existing studies of China's economic miracle?
Though much of what I want to say is in my referee report on Jessica Week’s paper, I want to reiterate here how much I enjoyed this article. I appreciate her attempt to argue that it is not just democratic leaders that face domestic punishment after foreign policy blunders such as losing a war. While I think there are various minor theoretical and methodological issues that should be resolved in further drafts, her work on identifying variation within autocracies that make some states more or less vulnerable to domestic pressure in foreign policy decision-making is interesting and fills an important gap in the international relations literature.
ReplyDeleteI also enjoyed Magaloni’s book, which was an interesting example of how elections, rather than indicating democratic consolidation, can serve to strengthen semi-authoritarian and authoritarian regimes already in place. This is certainly true in post-Soviet states, and Levitsky & Way (2002) have written some useful pieces that use façade elections to help them classify variation within authoritarian regimes. I think there are other mechanisms at work in the use of elections as a way to consolidate authoritarian rule that Magaloni doesn’t identify such as the fact that the elections can serve as an assertion of power of the populace, as the tactics used to render elections not fair and free can also emphasize the weakness and disenfranchisement of both the general population and the opposition. In demonstrating the weakness of the opposition, the regime underscores its position of power in society and legitimates itself both internationally and domestically. Further, by recognizing the international norm of elections as a pathway to democracy, the regime burnishes its own credentials and can use the rhetoric of democracy to achieve international standing and recognition as a “transitioning” state. More subtly, elections serve to strengthen semi-authoritarian regimes through the gauging of political support in a one and half party system and by strengthening patronage networks in the run-up to the elections Additionally, leaders can use the election has a means to institute un-free laws and institutions such as media and freedom of assembly laws. Even more nefarious techniques such as voter intimidation and increased police presence in opposition can also be used to institute an atmosphere of fear and trepidation in the run-up to an election. In contrast, leaders might also use the election time to appeal to the populace by starting long-overdue infrastructure projects or by handing out money and patronage. Another important aspect of elections in un-free states is that fact elections of this type are rarely issue based and instead become highly personalized and lacking in content. This type of electoral system removes politics from political issues and transforms the election into a personal advertisement for the autocrat and his regime. In these regimes, political parties are often centered around a small elite group of actors or a charismatic leader, and lack coherent ideologies. Since ideology has been removed from the election, the only constant in political life is the attitude toward the government, which is then confirmed by landslide election results in favor of the incumbent.
I find it interesting to compare and relationate the explanations given for the Mexican PRI's longetivity by both Haber and Magaloni (although I am definitely not a Mexicanist expert!). Whereas Haber emphasizes the importance of the logic of organizational proliferation, Magaloni's model (built on the Mexican case) gives a threefold arguments: behavioral incentives for elite unity, the strategic support of voters and the opposition's coordination dilemmas.
ReplyDeleteThe explanation of Haber is not extensive, but I believe it at least partly meets with the first and third dimensions of Magaloni's model. However, it seems to me that it does not recognize any importance for the electoral or even broader mass support to the autocracy. Neither does so the logic of terror described by Haber.
Personnally, I've been convinced by the importance of electoral/ mass support's argument of Magaloni, since the regime can't repress "so massively". One could qualify this support as falsified (as Kuran 1991, referred to by Magaloni, does it for Communist regimes); but I believe this does not diminish its importance as an explanation factor for the survival of autocracies/ hegemonic-party regimes.
I also enjoyed Magaloni’s work. While it was echoing a lot of Geddes, it answered a lot questions that arouse when reading the former, such as the ways in which elites and others are co-opted into the ruling party. Of her independent variables that explain how one-pary regimes survive and transition our of power—electoral support, electoral fraud, elite fragmentation and coordination challenges—I kept thinking about mass electoral support. One party regimes need a large turnout of voters and a large margin of victory at the polls to both deter challenges to their rule and the need to include any opposition groups that do exist. I started thinking about this but on a none-marco level. In the last legislative election in Lebanon, Hezbollah had the largest get out the vote drive, which included vote buying and intimidation, of all the parties even though their districts were largely uncontested for precisely the same reasons: they wanted to show that they represented the Shia voice in Lebanon, obviate any potential rivals from the traditional Shia families who resent for diluting their political power; and discount the need to incorporate other disparate views that lie within their districts, especially those of the communist party. At the same time, the party has established a good rapport and relationship with the other key Shia party, Amal, to the extent that they are considered to be co-opted. All these traits lend it to being a hegemonic one party actor among the Shia population of Lebanon. I also found the explanation to of why these regimes transition to democracy interesting and convincing.
ReplyDeleteI enjoyed Smith' paper on at least two counts:
ReplyDelete- First, I think it demonstrates the role for structured comparative case studies as a way of thinking about causality in small-N, comparative politics that is poorly captured by regressions (although the paper starts off with an all but meaningless replication of a 17-variable regime regression).
- Second, it allows for choices being made on different stages in a development: There are the initial conditions (struggle and rent), but there is also the question of what to do with rent if it appears later on, or how to deal with opposition later on.
However, there are also two things I don’t like:
- First, I don't think he elaborates enough on the possible theoretical strengths of this approach, and he is too eager to present clear narratives that fit the simplest variant of his theory - so simple, in fact, that it could be reduced to a regression (dependent variables would be long and hard civil war/struggle at the time of party strategy decision and access to rent). This seems unnecessary, since his cases presents a more complex picture: For instance, he claims that in Indonesia, the regime decided to use oil windfalls to deepen the party, rather than buying off people.
In terms of thinking about causality, such a two-step model, tracking mechanisms between initial conditions and outcome, is interesting. One could assume either that initial conditions, e.g. a deeper party structure increases the likelihood of subsequent rent to be used to strengthen the party (which Smith seems to do), or one could treat the relationship as random. Thinking of causality in terms of numerous steps connected through specified or unspecified probabilities might often more closely resemble how we think about causality, but often renders regression analysis uninterpretable. I think Smith’s narratives point to this, but it is not explicated.
- Second, the paper also shows some of the dangers of doing case-studies to confirm a theory: There seems to be a clear confirmation bias, i.e. in interpreting events in line with the theory. For instance, he seeks to portray PAIGC as a weak party, and seriously downplay the role of Portuguese forces in the country. First of all, although it is true that fighting was harder in other Portuguese territories, the opposition PAIGC met was arguably significantly stronger than what Nyerere encountered with the more peaceful British in Tanganyika. Second, Portuguese efforts increased over time, and in the late 1960s and early 1970s, they were pretty intense.
The bad choices made by Luiz Cabral are interpreted as excluding groups, whereas Nyerere’s are interpreted as including them – although there is probably some truth to this, it is striking that Nyerere’s dismissal of the existing elite (who were suspected of not being nationalistic) is not seen as exclusive, whereas groups excluded by Cabral are.
Likewise, Marco’s access to forestry revenues is seen as a huge boon, whereas Indonesian oil revenues prior to 1973 are just mentioned in passing as smaller than in 1960. By simplifying in this manner, I think Smith unnecessarily downplays the role of choices made by leaders at least partly independent from initial conditions.
I thought Kalyvas's article on the breakdown and decay of Soviet regimes was interesting, though perhaps his argument would be different today. While his point about the differences between breakdown and decay are well taken, the theory overall seems a bit dated, as today the variance between the former Soviet republics grows. Additionally, his comments on political competition no longer seem valid, as many of those who had been voted from power are now back in power, and if not former apparatchiks, at least their protege from the Komosomol. The description of decay and breakdown are thought-provoking, but do not they fall short because they do not entirely explain the reemergence of authoritarianism in much of the former Soviet Union, nor the diversity of outcomes. The patterns of decay and breakdown, while similar, must have some obvious differences that should be further detailed in order to make sense of current former post-Communist regimes.
ReplyDeleteI rather liked Weeks' paper, and I have to admit I was deeply envious of her army of research assistants who did all this wonderful archival work and coding. Her exposition was convincing, she frankly admitted the limitations of what she was doing, and she provided work arounds for possible problems in the coding.
ReplyDeleteNonetheless, I wonder if she couldn't have pushed her "war" variable a little more. Trimming the COW data to include only large or cross-border conflicts is all well and good, but perhaps there ought to be an additional distinction between limited and total wars, which could presumably be coded the same way as regime type (ie by using newspaper reports and academic accounts). Because it seems intuitively plausible that there are relevant differences between a minor border skirmish that didn't work out and losing World War II. Good meaures might be lost territory or some significant and costly concession.
Overall, a very well-done paper!
I limit myself to 3 articles because I realized my comments were becoming a bit long...
ReplyDeletea) Jessica L. Weeks
Here I have to be more critical than Robert. A critical point she doesn’t mention –as far as I remember- is that conventional wisdom is in fact true when we compare the ousting of autocratic leaders to democratic. It only doesn’t hold for “constrained” authoritarians. If I got this wrong, please correct me. If it is right, the paper’s contribution remains but is limited to a finer grain approach to autocracy, not to contradicting conventional wisdom.
An easy criticism –but not for that reason less valid, is that cross-tabulations are even more limited than regressions in determining causality. Therefore, in the paper we are given “empirical patterns” due to the small-n common problem. (Being a bit distrustful, could it be that regression results were not actually supporting her thesis?)
Finally, I have to agree with the author’s concern on page 35: “endogeneity between the leaders’ fates and the regime type codings is a real possibility”. In summary, this paper does not pretend to establish causality using methods like instrumental variables and instead builds a plausible argument with tentative results.
b) Benjamin Smith
It is a solid study that, as the author admits, needs to be tested. It reminded me of another comparative study by Clifford Geertz (Islam Observed, Religious Development in Morocco and Indonesia). I believe both authors make a good use of Mill’s method of agreement –that is, they compare disparate cases which have a commonality that is presumed the cause of the phenomenon. In Smith’s case, the cause would be the original strength of the party.
My critique is related to the graph on page 432: could something be more path-dependent? I understand/hope it is a simplification, and yet I can compare to game-theoretic assumptions that are just as little realistic.
c) Stephen Haber
I might be biased here because I very much enjoy the work of B. Weingast, not the least Violence and Social Orders. He uses technical concepts in a language understandable to undergraduates. I do think he reconciles the economics and sociology tradition simply by noting that the 3 winning strategies –which are 3 out of the 4 equilibria of the game- can be turned into categories. The 4th equilibrium is the dictator being ousted –certainly not a winning strategy from the dictator’s point of view…
I thought it was funny enough that Acemoglu et al. correct the econometric “misspecifications” of Przeworski once again, especially after reading Epstein’s et al. article in AJPS two weeks ago (see page 9).
A (minor?) critique to the paper (page 10) is whether the discontinuity factor (delta) grows with the dictator’s age. Is everyone like Mugabe? Or are some dictators less predatory as they grow older? That relates to a larger concern of the paper being very focused on the “predatory nature” of dictators, maybe ignoring that often times formal and informal constrains don’t allow them to be –like Tunisia nowadays.
Well, I can't resist making 2 short comments:
ReplyDeleteGehlbach and Keefer:
The authors develop a sophisticated model to for a rather intuitive (maybe obvious?) argument.
Schedler:
Good summary, but a boring article...
I was supposed to email everyone but I couldn't find a way of sending an email to the class on courseworks...anyway my referee report on the Weeks paper is up on courseworks in the shared files folder. @ joan in regards to your criticism of Weeks not using regression analysis: regressions are not always the most appropriate methodological choice in political science! I think she is right to use a cross-tab in this case, using a regression wouldn't capture the correlation she was looking for. I think there is a real bias in terms of quantitative analysis and in some cases being more explicit with the logic and less reliant on the numbers can be a good thing in my opinion. Second, Weeks does challenge conventional wisdom: the conventional wisdom is that it is the very nature of authoritarian regimes (lack of accountability to a domestic audience) that makes them immune to audience costs and less vulnerable to ouster. Here Weeks shows that this is not the case, some authoritarian regimes are vulnerable to ouster because they are more highly constrained by the selectorate. Instead of treating authoritarian regimes as a residual category, she builds on the large spate of recent literature that distinguishes between these regimes. Further, rather than just creating a typology of variation within authoritarian regimes she actually uses this difference to show a pattern of policy vulnerability. In any case, this is all just my opinion and I may be biased by the fact that I think she is doing great work on a really fascinating area of research (my own research area as it happens haha).
ReplyDelete@ Nils: Based on your comments, I would like to say something more about Smith. First of all, I am wondering if he can be more specific about the so-called "co-optation" through the establishment of a dominant party? Will there be any difference if the party is meant to co-opt the elite (as opposed to autocrats) or the mass? It will be surely good if Smith can focus something more on the process of party-building in these authoritarian regimes. Second, I think it is really hard to justify the selection of cases even though Smith does explain a little bit before he starts his case studies. How should we do this? I am also inclined to question if his selection is really appropriate. Thirdly, is Smith's article on the formation of dominant parties or the durability of dominant parties? This point is rather minor but may be worth discussing.
ReplyDeleteSome anecdotes about Mexican politics:
ReplyDeleteFirst I would like to say that from the literature related to the survival of the PRI I find that Magaloni builds a solid case for showing that elections were not devoid of content and its only function was to legitimize the regime. On the contrary elections were a very efficient information collection mechanism to be able to selectively coopt elites, union members and even the opposition.
In general I agree with the argument she presents however I would like to add some facts that I believe may help us understand the survival of the PRI better
1) While it is true that the populist wing and the neoliberal wing had different approaches as to how to handle the economy, it is also true that when it came to co-optation and clientelistic practices, both wings operated in the same way. For example, when Salinas (1988-1994) made explicit his intentions to reform, some union leaders were not willing to cooperate. Salinas' response was to incarcerate the leader of the oil workers union aka "la quina" and to change the leader of another important union the teachers. Therefore, signaling that he was the president and ultimately his decisions were law.
2) I don't think Magaloni gives sufficient attention to the effects that signing NAFTA had on the Mexican system. Signing NAFTA required demonstrating that the country could at least have stable macroeconomic indicators thus, by the time Salinas was trying to enter into this treaty, he had significant pressure from abroad, thus restraining to some extent the policies he could explicitly carry.
3) I am not sure if a regression is necessary to demonstrate that PRONASOL was in fact an efficient instrument for vote buying. I believe that the evidence presented in a case study is sufficient for proving that PRONASOL was designed to buy back defectors to the PRD. For instance, I would add the fact that Salinas' presidential candidate was the Minister in charge of administering this program. In addition, presidents had "reserved funds" that they also used to keep their co optation system.
In line with Magaloni I also provide some phrases from the Mexican politicians that I believe have great explanatory power:
"I don't see them, and I don't hear them" Carlos Salinas in reference to the opposition claiming fraud in the 1988 elections.
"I will defend the peso like a dog" Jose Lopez Portillo-the next day the peso was devaluated…
"If you think it don't say it, if you say it don't sign it, and if you sign it back off" Fidel Velazquez Union Leader
@ Joan: I think your points on Week's article are very interesting. Even though I am not an expert on regression analysis, I think it will be a nice try if you can do this. I do not think there is much to say which one is good before we see the results, but there is still a chance to tell which one may better support her argument. Also, yeah, I think her point in this article is not to contrast autocracy and democracy but to dissect the concept "autocracy" and to explore the variation within this category.
ReplyDeleteBy the way, I totally agree with your comment on Schedler's article! While it is very straightforward and clear, I really think it can be arranged in a more readable way.
I really liked this week's readings, but something struck me as I was doing the readings: there is hardly any mention of the possible role of the bureaucracy in supporting authoritarian regimes in this week's readings. Although Smith has provided a cogent argument and strong evidence through 4 case studies that the origins rather than the existence of single-party rule are crucial in determining the viability of such regimes, there is no mention of roles that bureaucratic agents might have played in reinforcing single-party rule. However, the possibly significant role that bureaucracies might have played is hinted at by Haber in his paper, where he notes that "the dictator needs this launching organization, as well as the other organizations within the state that the launching group permeates, because without them he cannot run the country". Indeed, one could argue that any dictator or ruling party would necessarily depend upon bureaucratic agents with specific areas of expertise to aid them in their rule. Hence it might be interesting to consider how the relationship between the single-ruling party and the bureaucracy in authoritarian regimes could have possibly reinforced the viability of the ruling-party, e.g. whether the ruling party has awarded significant benefits to the bureaucrats to incentivise them to work for the party; whether bureaucratic drift could eventually lead the ruling party to a downfall because the ruling party alone lacks the necessary capabilities and expertise to run all aspects of the country.
ReplyDelete