Many readings of this week aim to explore "how institutions affect economic growth." Even though Jones and Olken's argue that the (natural) death of national leaders has large impact on economic development of a country, they also find out such effect is still conditioned by the institutional context (democracy or autocracy). In terms of methods, all of the researchers heavily rely on quantitative data and statistical techniques to confirm their theoretical arguments, but Stasavage, North et al., and Glaeser et al. in the meanwhile also consult history to test whether their theoretical claims truly correspond to the historical evidence. While I really enjoy reading all of them, I still have some concerns about their approaches and the way they prove their argument in each article.
To begin with, all authors do justify the use of data and models in their work, which is surely compelling, but at times I still cannot see the causal mechanism very clearly. For instance, Jones and Olken, as economists, apply quite sophisticated quantitative analysis that let them find out highly tenuous relationship between the death of national leader and the growth of national economy. However, what really causes such effect? Why does a leader's death, especially the ones in autocracy, tend to lead to economic growth? Can the economy grow even when the leaders are still alive? The coefficients cannot inform us of these nuances, and I expect them to elaborate more since I still doubt the legitimacy regarding the use of dichotomous category (democracy vs. autocracy). For instance, they use Taiwan as an example, but most scholars do not consider the death of two Chiangs as the driving force of economic growth in Taiwan since it is really the bureaucracy leading the process of economic decision-making. I understand economists prefer parsimony, but as a student of political science, I do hope to see more case studies to reveal the mechanism hidden behind the numbers. Second, perhaps more importantly, I notice that many scholars tend to presume the existence of the causal connection between institutions and economic development. Take North and Weingast as the example, they argue that the establishment of parliamentary supremacy and multiple veto-points in the English political system helped the accumulation of capital that further encouraged the industrialized development. However, they prove their argument simply by presenting the facts of institutional change and the changes in the interest rates of government loans. Doing so may tempt the readers to buy their argument, but I still feel uneasy about this practice especially because this practice is also present in other articles. So the question is: How can we really ascertain the causal stories we would like to delineate? Self-falsification (e.g. discussing how competing argument relates to our own work) or adding one more case (e.g. Glaeser) may be a good beginning, but is there any other way to alleviate this problem?
Therefore, while I know the connection between institutions and economic development is indeed too robust to be ignored, I still hesitate to say I have learned very much about how it is the case. Do institutions foster economic prosperity only/mainly because they induce the stability of the exercise of property rights? More crucially, why do institutions turn out to be the strongest independent variable? Regarding this matter, I think Rodrik et al. have done a really interesting and excellent job even though their results demand some more case/empirical studies.
For individual readings, although I really think Stasavage has done an excellent job to elaborates North and Wiengast's work, his reference of Persson and Tabellini's research is a little bit confusing. It seems like it is Figure 6 instead of the literature that motivates his article. I understand he may want to use PT as the inspiration to tackle the connection between redistribution and partisan politics, but it may not be needed as he could just identify how partisan politics relate to the fluctuation of interest rates after 1688.
For Glaeser et al., their study on the corporate law and the securities law in Poland and Czech Republic is definitely amazing. However, I wish they could explain more on how these two laws relate to each other. It seems like they care more about the latter (as the Polish law on securities regulation succeeds in reducing the value of c and increasing the value of a). There is only one line regarding the relationship between the two in p.887. By the way, their discussion on Hungary appears to be rather strange (especially because there is no equally extensive information provided). Does adding one more case help strengthen their argument?
Last but not least, some China scholars do consult North and Weingast to explore the post-1980 economic development (e.g. Jean Oi). It is very interesting that they argue the property rights in China is not protected based on individuals but based on groups, namely local governments. As the central government decentralizes the fiscal system and delegate the management of many SOEs to the local cadres, who now are free to retain and spend the revenues they earn, the Chinese economy takes off.
Oi, Jean C. 1992. "Fiscal Reform and the Economic Foundations of Local State Corporatism in China," World Politics 45 (1): 99-126.
Chao-yo, in regards to Olken and Jones, I was under the impression that (although not explicitly stated), the mechanism behind growth was the leader's ability to direct institutional changes? The examples of Mao and Khomeini suggest that leaders institute a variety of policies that may affect growth. I assume that this effect is less pronounced for democratic regimes, because it is more difficult for one person to influence policy outcomes. Regardless, it's an issue that needs further clarification.
I really enjoyed the OJ article, and thought it brought to light some interesting issues regarding the role of leaders. I wish there had been more theory about how leaders influence (or not) growth. For instance, while leader deaths seem to influence growth in some direction, it seems counterintuitive to me to expect leaders' deaths to influence growth in autocratic regimes where a succession plan is in place. Are people more optimistic than myself about new leaders? I would expect a new leader to generally carry out similar policies to the former leader, especially in certain types of regimes (um, present-day Russia, between Medvedev and Putin?)...Like Chao-yo, I am skeptical of the use of dichotomous variables.
For this weeek I would like to comment about the article Coase Versus The Coasians by Glaeser, Johnson and Shleifer. First, I find refreshing that they include payoffs in the form of self-esteem or long-run respect of the peers -the b in the model- that confirms the work by Chester I. Barnard who argues that “opportunities for distinction, prestige, personal power, and the attainment of dominating position are much more important than material rewards in the development of all sorts of organizations.” The inclusion of such non-material considerations, even though harder for building a case empirically, yields a superior explanatory power.
However I have two questions about their model that I would like to present. 1) They argue that extensive disclosure of financial and ownership information lowers the costs of gathering the information and thus faciliates regulation. Nonetheless, it is not clear to me that more transparency necessarily leads to lowering costs inasmuch we should also consider how the information is presented. That is, transparency is not only relevant but most importantly how the information is disclosed.
2)The authors conclude by saying that the regulated Polish stock market has been a better source of capital than the less regulated Czech market (892). The important question, however, is how much regulation is good regulation. COnsequently it is a matter of degree and how this interplays with North's informal institutions is perhaps worth exploring.
About the article by Stasavage, it is not clear to me how much his revisiting of the North and Weingast argument is a clear contribution to understanding how securing property rights developed institutions. In my view, North and Weingast explain in its majority the development in England of private credit and economic growth and Stasavage's argument is just marginal in its contribution and not equally important as he claims it to be.
A question for STAT LOVERS:
In the Jones and Olken data, they have leaders that were in power in different periods, for example Peron in Argentina. Could this change their results? That is, if what they are trying to explain is how the death of a leader affects economic growth, could it be argued that by including in your data leaders who have been in power in different moments this changes your results?
Kim: I agree with your contention about leader death’s influencing growth in autocratic regimes with succession plans but even though there are a few prominent exceptions, the fact remains that any autocratic leadership transitions are times of great uncertainty in the regime. However, I think what you say is true and different types of autocratic regimes would probably vary greatly on this extent (leader death in sultanistic regimes would probably be much more likely to influence growth when say compared to military dictatorship where policies are more institutionalized) to which leader death would influence growth.
In any case, my comments this week are brief: first, I enjoyed the JO piece but found it amusing that they used the scene where Napoleon is feeling under the weather because of wet boots in War and Peace as evidence that Tolstoy was dismissive of the effects of leadership in their literature review. Like he did on many other subjects, Tolstoy vacillated throughout his writings on his views about the power of leaderships and historical determinism; I think the article would have been more powerful had JO used Burns in their review to set up the paper.
Second, I also enjoyed the GJS article but had a few substantive comments. Although I’m not sure what impact (if any) this would have on their study, I was troubled by the fact that JO claim that Poland and the Czech Republic and to a lesser extent Hungary had similar initial conditions. In economic terms, the Czech Republic had one of the lowest hard currency debts in Eastern Europe, by contrast Poland had accrued a huge debt to Western Europe and experienced hyperinflation. Additionally, Hungary was also massively in debt to the West and would have faced insolvency several times had they not been able to secure Western loans.
In Unbundling Institutions, Acemolgu and Johnson build upon their argument that the institutional legacy of colonialism affects economic growth by concluding that property right institutions have a greater influence on economic growth (per income) and on the likelihood of investment than contracting institutions, which primarily influence financial aspects of the economy like the stock market. I found this interesting, though not surprising given the role property rights have played in political development. Specifically, the role it played during the Britain’s path to democracy and its role in the French Revolution both come to mind. Yet in thinking about these cases, there already existed a belief that land can be privately owned. Yes colonial legacy did transplant some of these institutions to parts of the world where such a concept did not exist but has the concept really taken root to the fullest extent? I ask this with certain cases in the Middle East in mind where the concept of private property exists but its role differs: instead of acting as a constraining mechanism against government they are used to by government to consolidate power.
1) I very much liked the way Rodrik et al. 2004 acknowledge the difference relations between the determinants of income often opposed in the literature (geography, integration, institutions). However, I think their operationalization does not adequately represent their theoretical framework. For example, they use the AJR measure of settler mortality to measure institutions, although I think settler mortality, especially since AJR estimated great parts of their on the basis of disease environment, is greatly linked to geography. If an instrument variable is linked to to variable, how can they use it to confirm the primary impact of institutions, over geography? Maybe my reading misunderstood part of their statistical analysis, and if so, I hope someone can clarify this to me...
2) That's a more general comment I may have already done, but I am still stunned by the importance given to the explanation of the differences in GDP per capita. I think differences in terms of inequality would be more interesting to explain. And even if not going to inequality, could we not look at a measure analogous to the GDP per capita that could take into account the differences in living expenses in the different countries? I think it would, this way, be more relevant and useful.
3. I thought the historical detailled perspective of North and Weingast was very interesting and clear, and informative of a period of English history I know very little about. However, their very brief comparison with France at the very end, responding to some literature, was not clear to me. I thought their argument then lost of its strength. As such, I think Stasavage completes very well their work and bonifies it in an interesting way, by going beyond the specific English case.
I am quite a fan of Weingast and North, so I might not be very objective. The theory is very “compelling” (Stasavage dixit) even if tests are difficult. The main broad reaching idea is that expectations play a key role in the degree of investment (and trust), it is about predictability and commitment. Interestingly, this leads to the question of whether democracy can be reduced to a normative preference. However, this question is intendedly spurious. That is, if Singapore respects property rights and the like, why do we need democracy? Simply, because a democratic setting minimizes the imperfection in the aggregation of preferences. If you believe the rights to trade and buy are as important (economic rights) as positive and negative liberty (civil and political rights), then the case for democracy is as compelling as the case for reliable institutions that allow optimal economic performance.
I believe Glaeser et al. present a compelling case, even if causality in these instances is always arguable (see their own “VII. Discussion” Section). Ironically, it seems that Vaclav Klaus’ “passion” for a supposedly rational and homo economicus ideology (U. of Chicago) performed poorly compared to an actually rational and thought out approach by the Polish regulator Balcerowicz in which he tries to determine the extent to which state capacity should be exercised.
Coase’s Theorem rests on two premises (Glaeser et al, p. 853): well-defined property rights and no transactions costs. Since the 2nd assumption is never met, like Coase admits, the issue at stake is how to deal with inefficiency. The key for Coase is a cost-benefit analysis, which involves the quantification of the transaction costs. The normative question is: should a comparison of cost and benefit be the criterion? It being the only criterion might create a large social cost in itself. A real example follows. Back home, my mother is bothered in the evening by the three little kids of the neighbors upstairs, who are very loud. She went up to complain and the noise was reduced, however to a substantially lesser extent that she would have desired. Why didn’t she push it further? Because bringing it up to the community president would result in damaging social relations in the building by creating two sides: those in favor of my mother and those in favor of the neighbors. Is the cost-benefit analysis broad enough when we deal with non-monetary issues?
Regarding Jones and Olken, a critical question is the quality of the IV. They correctly eliminate assassinations, leaving 65+12=77 cases, but they can only use 57 given data constrains (page 31). It is legitimate to ask whether this leads to bias. A big triumph in their analysis is that the 57 cases are not significantly different –except for age- than the over 1,184 total observations. Another point is the division of the Polity score turning scores below 0 into autocracy and above 0 into democracy. This is arguably a very low threshold. It seems that last year a group replicated the results of a paper that fell apart for the most part by changing this threshold. This is not a main variable in that paper, however. Finally, is the PRIO database a good proxy for changes in security policy? Besides these minor objections, I believe it is a solid and even unconventional paper.
I would like to comment on the Jones & Olken this week.
I agree with Kim that there ought to be more theory in the paper on how leaders actually influence growth. Echoing Kim's comments, I believe that the authors' implicit causal link between leaders and growth is that some leaders, e.g. Mao, have the capability to establish and preserve a set of unique institutions during their reign, but these institutions are prone to give way to other institutions with leadership transition. For instance, the authors gave the example of China, where the poor growth that was observed during Mao's tenure could most likely be attributed to policies e.g. collectivization of agriculture, Cultural Revolution, that Mao had adopted. On the other hand, China experienced much faster growth under Deng Xiaoping and the 1979 economic reforms that he had instituted.
My main problem and question with their analysis is: What, then, is the difference between these leaders and the institutions/policies that they have implemented? If a leader can be characterized by his policies, then does it actually matter whether we study leaders and leadership transition, or if we simply focus on the policies themselves? e.g. Why should we study Mao and what happened to China's growth before and after his tenure if we could instead focus on socialism and socialist policies (such as collectivization of agriculture) as ideological and institutional constraints on growth?
The true causal story might not be that leaders matter for growth, but that institutions matter for growth, and leaders are important insofar as they precipitate or prevent institutional change. Indeed, the authors hinted on this in their comparison of the impact of leadership transitions in autocratic versus democratic regimes. They state that "one might expect that the degree to which leaders can affect growth depends on the amount of power vested in the national leader"; in other words, the power of the leader during his tenure and subsequently, the path dependency and institutional inertia associated with the programs implemented during his tenure. Not only does this require further analysis of the institutional framework and rules that influence the extent of path dependency, this also requires further elaboration on the role of the bureaucracy or the political parties that have aided the leader in implementing his policies. e.g. Mao could not have governed China on his own and had to rely heavily on the Chinese Communist Party for support. It would certainly be interesting for the authors to explore questions such as the interactions between the leader and his political party or the bureaucracy to explain how institutional changes could arise within the same party, albeit led by different leaders.
Interestingly, leaders could matter for growth beyond their tenure even if there is no significant difference in growth before and after the leader's tenure. The authors mentioned "the successful economic policies of Lee Kuan Yew in Singapore". Indeed, political leaders in Singapore have invested heavily and continuously in grooming subsequent generations of leaders as well as the bureaucracy to ensure that political instability would be minimized in the event of a leadership transition and that economic growth would be sustained. In this instance, the lack of any difference in growth rates before and after a leadership transition does not necessarily imply that leaders matter less for growth; here the impact of leaders simply works through a different avenue.
Again, not really an "Econ" person, but really enjoy the readings, Coase in particular. I suppose this is because I began with my standard eye rolling about the idea that that rancher or farmer could know all this information, think about it in those specific terms, and engage in a negotiation in the manner described. But then, BOOM, he brings in the cost of information, which maps really well onto my lived experience, where finding the right price and doing the appropriate inspections of a large purchase are, at a minimum, very time consuming and often monetarily costly as well.
@ Kim and Man Yan: I learn a lot from your comments on Jones & Olken's article. After reading the article again by myself, I, too, am puzzled by the true mechanism between leaders and growth. While it is tempting to argue that the Chinese economy took off during the 1980s because Mao passed away, I will say China made it because the leadership chose the right policies, which was surely not an easy thing because the Chinese Communist Party suffered greatly internal struggle between the conservative faction (led by Chen Yun) and the reform faction (headed by Deng). Therefore, what we can imagine is if Deng failed to persuade the whole party leadership to adopt market economy, then the post-1980 Chinese history will be totally different. And Man Yan is surely right that the influence of a leader will continue even after he is gone, and there is still variation across leaders regarding their impact and control over policy-decision making that cannot be detected simply based on the difference between democracy and autocracy. So I guess two conclusions I have here are: One, it may be dangerous to give the leader too much credit. Two, democratic leaders may not necessarily weaker or less influential compared to their autocratic colleagues. I understand Jones and Olken did recognize these nuances, but I am still not satisfied with their results because I think there is simply more to say beyond statical results.
A few thoughts on the North and Weingast article and the ensuing debate:
- OK, this is a minor side track, but the way in which rational choice is applied in trying to explain historical developments sometimes annoys me. Page 807: “In many of the simple repeated games studied in the literature, this incentive alone is sufficient to prevent reneging” – followed by some “rational” exceptions to this rule, in the example of the confiscation by Edward I of Jewish property in 1290. However, the explanation seems to me a post hoc rationalization. More immediate explanatory variables spring to mind: There was increasing anti-semitism in England since the resurgence of the blood libel in the 12th century, and also complaints about usury. Jews in England were the Sovereign’s personal property, and they could be taxed at will. An important reason why Jews mattered less economically at the time of confiscation in 1290 was that Edward I had already taxed them so heavily, so that there was not much left. Veight, whom North and Weingast cite, provides no evidence that Edward I in actuality considered the relative value of Jews against Italians; nor that he felt that the economic situation was such that he no longer needed them. In fact, Edward I constantly sought funds from any source he could find, and he used the expulsion of Jews, which was sought by Parliament, as an argument for getting more taxes from the English. There is every reason to believe that anti-semitism on the part of Parliament, probably Edward I (who, however, also seems to have felt some pressure from Parliament in this matter) played the major role of the decisions in this regard, leading to expulsion in 1304. Phillip II of France expelled Jews from crown lands, in 1192, John I of Brettany expelled them in 1239, and a little later Louis IX of France expelled them from France. All of them experienced revenue loss. It is possible to construct games where this course of action is rendered rational, but it transforms the concept of rationality into a truism without analytic or substantive meaning.
- I find North and Weingast’s general point to be relevant, and Stasavage’s reservations and elaborations to be generally useful. We know from many concrete situations that a parliament is not seen as a guarantee against expropriation. Indeed, in interwar Europe, the fair of radical, expropriationist parliaments abounded in Europe, and contributed to capital flight, class struggle and fascism. It is not unlikely that this was also the case in England and France in the 18th century.
However, it is difficult to say with any confidence from Stasavage’s account. The main evidence that he in my view presents, in addition to theoretical considerations, is the spike in interest rates during unified Tory rule from 1710 onwards. However, this period coincides with the last phase of the War of the Spanish Succesion. The costs of that war drove interest rates up everywhere. Unified Tory rule might have contributed, but we do not know. Whether London financiers and landowners behaved, or were believed to behave, differently especially in the English Parliament, Stasavage investigates to superficially. However, there is a huge debate about a similar question in the US Federal Convention. In general, it has been difficult to establish the effect of economic interests in this regard. There might be ideological or religious questions involved (repay your debt), but there are also other, more vague questions of interest involved. Bear in mind that we at any rate discuss a wealthy elite, though their sources of wealth differed. Arbitrariness was welcomed by neither financier nor landowner. The popular legislatures of the American colonies, on the other hand, with representation not only from the elites, but from the if not “the common man”, then at least people without personal wealth, tended to be less keen on repaying government debts.
- An example which lends contemporary relevance to the mechanism proposed by North and Weingast can be found on FrumForum today:
"Some thoughts on the country he led, inspired by a recent visit:
Here’s a story I heard last week from a onetime foreign investor in Argentina. The thing that drove him out of the country was a 5 percent tax on his company. Five percent may not sound like much, but what mattered was not the amount of the tax. It was the way it was imposed. The tax was not enacted by Congress. It was not even ordered by the president.
One fine day, one of his operations received a letter from a government ministry with a sudden demand for payment. As far as he could tell, the demand was ungrounded in any law. He litigated the matter and (eventually) prevailed. But a country where the government could remake the rules at any time was no country for him."
Many readings of this week aim to explore "how institutions affect economic growth." Even though Jones and Olken's argue that the (natural) death of national leaders has large impact on economic development of a country, they also find out such effect is still conditioned by the institutional context (democracy or autocracy). In terms of methods, all of the researchers heavily rely on quantitative data and statistical techniques to confirm their theoretical arguments, but Stasavage, North et al., and Glaeser et al. in the meanwhile also consult history to test whether their theoretical claims truly correspond to the historical evidence. While I really enjoy reading all of them, I still have some concerns about their approaches and the way they prove their argument in each article.
ReplyDeleteTo begin with, all authors do justify the use of data and models in their work, which is surely compelling, but at times I still cannot see the causal mechanism very clearly. For instance, Jones and Olken, as economists, apply quite sophisticated quantitative analysis that let them find out highly tenuous relationship between the death of national leader and the growth of national economy. However, what really causes such effect? Why does a leader's death, especially the ones in autocracy, tend to lead to economic growth? Can the economy grow even when the leaders are still alive? The coefficients cannot inform us of these nuances, and I expect them to elaborate more since I still doubt the legitimacy regarding the use of dichotomous category (democracy vs. autocracy). For instance, they use Taiwan as an example, but most scholars do not consider the death of two Chiangs as the driving force of economic growth in Taiwan since it is really the bureaucracy leading the process of economic decision-making. I understand economists prefer parsimony, but as a student of political science, I do hope to see more case studies to reveal the mechanism hidden behind the numbers. Second, perhaps more importantly, I notice that many scholars tend to presume the existence of the causal connection between institutions and economic development. Take North and Weingast as the example, they argue that the establishment of parliamentary supremacy and multiple veto-points in the English political system helped the accumulation of capital that further encouraged the industrialized development. However, they prove their argument simply by presenting the facts of institutional change and the changes in the interest rates of government loans. Doing so may tempt the readers to buy their argument, but I still feel uneasy about this practice especially because this practice is also present in other articles. So the question is: How can we really ascertain the causal stories we would like to delineate? Self-falsification (e.g. discussing how competing argument relates to our own work) or adding one more case (e.g. Glaeser) may be a good beginning, but is there any other way to alleviate this problem?
Therefore, while I know the connection between institutions and economic development is indeed too robust to be ignored, I still hesitate to say I have learned very much about how it is the case. Do institutions foster economic prosperity only/mainly because they induce the stability of the exercise of property rights? More crucially, why do institutions turn out to be the strongest independent variable? Regarding this matter, I think Rodrik et al. have done a really interesting and excellent job even though their results demand some more case/empirical studies.
For individual readings, although I really think Stasavage has done an excellent job to elaborates North and Wiengast's work, his reference of Persson and Tabellini's research is a little bit confusing. It seems like it is Figure 6 instead of the literature that motivates his article. I understand he may want to use PT as the inspiration to tackle the connection between redistribution and partisan politics, but it may not be needed as he could just identify how partisan politics relate to the fluctuation of interest rates after 1688.
ReplyDeleteFor Glaeser et al., their study on the corporate law and the securities law in Poland and Czech Republic is definitely amazing. However, I wish they could explain more on how these two laws relate to each other. It seems like they care more about the latter (as the Polish law on securities regulation succeeds in reducing the value of c and increasing the value of a). There is only one line regarding the relationship between the two in p.887. By the way, their discussion on Hungary appears to be rather strange (especially because there is no equally extensive information provided). Does adding one more case help strengthen their argument?
Last but not least, some China scholars do consult North and Weingast to explore the post-1980 economic development (e.g. Jean Oi). It is very interesting that they argue the property rights in China is not protected based on individuals but based on groups, namely local governments. As the central government decentralizes the fiscal system and delegate the management of many SOEs to the local cadres, who now are free to retain and spend the revenues they earn, the Chinese economy takes off.
Oi, Jean C. 1992. "Fiscal Reform and the Economic Foundations of Local State Corporatism in China," World Politics 45 (1): 99-126.
Chao-yo, in regards to Olken and Jones, I was under the impression that (although not explicitly stated), the mechanism behind growth was the leader's ability to direct institutional changes? The examples of Mao and Khomeini suggest that leaders institute a variety of policies that may affect growth. I assume that this effect is less pronounced for democratic regimes, because it is more difficult for one person to influence policy outcomes. Regardless, it's an issue that needs further clarification.
ReplyDeleteI really enjoyed the OJ article, and thought it brought to light some interesting issues regarding the role of leaders. I wish there had been more theory about how leaders influence (or not) growth. For instance, while leader deaths seem to influence growth in some direction, it seems counterintuitive to me to expect leaders' deaths to influence growth in autocratic regimes where a succession plan is in place. Are people more optimistic than myself about new leaders? I would expect a new leader to generally carry out similar policies to the former leader, especially in certain types of regimes (um, present-day Russia, between Medvedev and Putin?)...Like Chao-yo, I am skeptical of the use of dichotomous variables.
For this weeek I would like to comment about the article Coase Versus The Coasians by Glaeser, Johnson and Shleifer.
ReplyDeleteFirst, I find refreshing that they include payoffs in the form of self-esteem or long-run respect of the peers -the b in the model- that confirms the work by Chester I. Barnard who argues that “opportunities for distinction, prestige, personal power, and the attainment of dominating position are much more important than material rewards in the development of all sorts of organizations.” The inclusion of such non-material considerations, even though harder for building a case empirically, yields a superior explanatory power.
However I have two questions about their model that I would like to present.
1) They argue that extensive disclosure of financial and ownership information lowers the costs of gathering the information and thus faciliates regulation. Nonetheless, it is not clear to me that more transparency necessarily leads to lowering costs inasmuch we should also consider how the information is presented. That is, transparency is not only relevant but most importantly how the information is disclosed.
2)The authors conclude by saying that the regulated Polish stock market has been a better source of capital than the less regulated Czech market (892). The important question, however, is how much regulation is good regulation. COnsequently it is a matter of degree and how this interplays with North's informal institutions is perhaps worth exploring.
About the article by Stasavage, it is not clear to me how much his revisiting of the North and Weingast argument is a clear contribution to understanding how securing property rights developed institutions. In my view, North and Weingast explain in its majority the development in England of private credit and economic growth and Stasavage's argument is just marginal in its contribution and not equally important as he claims it to be.
A question for STAT LOVERS:
In the Jones and Olken data, they have leaders that were in power in different periods, for example Peron in Argentina.
Could this change their results? That is, if what they are trying to explain is how the death of a leader affects economic growth, could it be argued that by including in your data leaders who have been in power in different moments this changes your results?
Kim: I agree with your contention about leader death’s influencing growth in autocratic regimes with succession plans but even though there are a few prominent exceptions, the fact remains that any autocratic leadership transitions are times of great uncertainty in the regime. However, I think what you say is true and different types of autocratic regimes would probably vary greatly on this extent (leader death in sultanistic regimes would probably be much more likely to influence growth when say compared to military dictatorship where policies are more institutionalized) to which leader death would influence growth.
ReplyDeleteIn any case, my comments this week are brief: first, I enjoyed the JO piece but found it amusing that they used the scene where Napoleon is feeling under the weather because of wet boots in War and Peace as evidence that Tolstoy was dismissive of the effects of leadership in their literature review. Like he did on many other subjects, Tolstoy vacillated throughout his writings on his views about the power of leaderships and historical determinism; I think the article would have been more powerful had JO used Burns in their review to set up the paper.
Second, I also enjoyed the GJS article but had a few substantive comments. Although I’m not sure what impact (if any) this would have on their study, I was troubled by the fact that JO claim that Poland and the Czech Republic and to a lesser extent Hungary had similar initial conditions. In economic terms, the Czech Republic had one of the lowest hard currency debts in Eastern Europe, by contrast Poland had accrued a huge debt to Western Europe and experienced hyperinflation. Additionally, Hungary was also massively in debt to the West and would have faced insolvency several times had they not been able to secure Western loans.
In Unbundling Institutions, Acemolgu and Johnson build upon their argument that the institutional legacy of colonialism affects economic growth by concluding that property right institutions have a greater influence on economic growth (per income) and on the likelihood of investment than contracting institutions, which primarily influence financial aspects of the economy like the stock market. I found this interesting, though not surprising given the role property rights have played in political development. Specifically, the role it played during the Britain’s path to democracy and its role in the French Revolution both come to mind. Yet in thinking about these cases, there already existed a belief that land can be privately owned. Yes colonial legacy did transplant some of these institutions to parts of the world where such a concept did not exist but has the concept really taken root to the fullest extent? I ask this with certain cases in the Middle East in mind where the concept of private property exists but its role differs: instead of acting as a constraining mechanism against government they are used to by government to consolidate power.
ReplyDelete1) I very much liked the way Rodrik et al. 2004 acknowledge the difference relations between the determinants of income often opposed in the literature (geography, integration, institutions). However, I think their operationalization does not adequately represent their theoretical framework. For example, they use the AJR measure of settler mortality to measure institutions, although I think settler mortality, especially since AJR estimated great parts of their on the basis of disease environment, is greatly linked to geography. If an instrument variable is linked to to variable, how can they use it to confirm the primary impact of institutions, over geography? Maybe my reading misunderstood part of their statistical analysis, and if so, I hope someone can clarify this to me...
ReplyDelete2) That's a more general comment I may have already done, but I am still stunned by the importance given to the explanation of the differences in GDP per capita. I think differences in terms of inequality would be more interesting to explain. And even if not going to inequality, could we not look at a measure analogous to the GDP per capita that could take into account the differences in living expenses in the different countries? I think it would, this way, be more relevant and useful.
3. I thought the historical detailled perspective of North and Weingast was very interesting and clear, and informative of a period of English history I know very little about. However, their very brief comparison with France at the very end, responding to some literature, was not clear to me. I thought their argument then lost of its strength. As such, I think Stasavage completes very well their work and bonifies it in an interesting way, by going beyond the specific English case.
I am quite a fan of Weingast and North, so I might not be very objective. The theory is very “compelling” (Stasavage dixit) even if tests are difficult. The main broad reaching idea is that expectations play a key role in the degree of investment (and trust), it is about predictability and commitment. Interestingly, this leads to the question of whether democracy can be reduced to a normative preference. However, this question is intendedly spurious. That is, if Singapore respects property rights and the like, why do we need democracy? Simply, because a democratic setting minimizes the imperfection in the aggregation of preferences. If you believe the rights to trade and buy are as important (economic rights) as positive and negative liberty (civil and political rights), then the case for democracy is as compelling as the case for reliable institutions that allow optimal economic performance.
ReplyDeleteI believe Glaeser et al. present a compelling case, even if causality in these instances is always arguable (see their own “VII. Discussion” Section). Ironically, it seems that Vaclav Klaus’ “passion” for a supposedly rational and homo economicus ideology (U. of Chicago) performed poorly compared to an actually rational and thought out approach by the Polish regulator Balcerowicz in which he tries to determine the extent to which state capacity should be exercised.
Coase’s Theorem rests on two premises (Glaeser et al, p. 853): well-defined property rights and no transactions costs. Since the 2nd assumption is never met, like Coase admits, the issue at stake is how to deal with inefficiency. The key for Coase is a cost-benefit analysis, which involves the quantification of the transaction costs. The normative question is: should a comparison of cost and benefit be the criterion? It being the only criterion might create a large social cost in itself. A real example follows. Back home, my mother is bothered in the evening by the three little kids of the neighbors upstairs, who are very loud. She went up to complain and the noise was reduced, however to a substantially lesser extent that she would have desired. Why didn’t she push it further? Because bringing it up to the community president would result in damaging social relations in the building by creating two sides: those in favor of my mother and those in favor of the neighbors. Is the cost-benefit analysis broad enough when we deal with non-monetary issues?
Regarding Jones and Olken, a critical question is the quality of the IV. They correctly eliminate assassinations, leaving 65+12=77 cases, but they can only use 57 given data constrains (page 31). It is legitimate to ask whether this leads to bias. A big triumph in their analysis is that the 57 cases are not significantly different –except for age- than the over 1,184 total observations. Another point is the division of the Polity score turning scores below 0 into autocracy and above 0 into democracy. This is arguably a very low threshold. It seems that last year a group replicated the results of a paper that fell apart for the most part by changing this threshold. This is not a main variable in that paper, however. Finally, is the PRIO database a good proxy for changes in security policy? Besides these minor objections, I believe it is a solid and even unconventional paper.
I would like to comment on the Jones & Olken this week.
ReplyDeleteI agree with Kim that there ought to be more theory in the paper on how leaders actually influence growth. Echoing Kim's comments, I believe that the authors' implicit causal link between leaders and growth is that some leaders, e.g. Mao, have the capability to establish and preserve a set of unique institutions during their reign, but these institutions are prone to give way to other institutions with leadership transition. For instance, the authors gave the example of China, where the poor growth that was observed during Mao's tenure could most likely be attributed to policies e.g. collectivization of agriculture, Cultural Revolution, that Mao had adopted. On the other hand, China experienced much faster growth under Deng Xiaoping and the 1979 economic reforms that he had instituted.
My main problem and question with their analysis is: What, then, is the difference between these leaders and the institutions/policies that they have implemented? If a leader can be characterized by his policies, then does it actually matter whether we study leaders and leadership transition, or if we simply focus on the policies themselves? e.g. Why should we study Mao and what happened to China's growth before and after his tenure if we could instead focus on socialism and socialist policies (such as collectivization of agriculture) as ideological and institutional constraints on growth?
The true causal story might not be that leaders matter for growth, but that institutions matter for growth, and leaders are important insofar as they precipitate or prevent institutional change. Indeed, the authors hinted on this in their comparison of the impact of leadership transitions in autocratic versus democratic regimes. They state that "one might expect that the degree to which leaders can affect growth depends on the amount of power vested in the national leader"; in other words, the power of the leader during his tenure and subsequently, the path dependency and institutional inertia associated with the programs implemented during his tenure. Not only does this require further analysis of the institutional framework and rules that influence the extent of path dependency, this also requires further elaboration on the role of the bureaucracy or the political parties that have aided the leader in implementing his policies. e.g. Mao could not have governed China on his own and had to rely heavily on the Chinese Communist Party for support. It would certainly be interesting for the authors to explore questions such as the interactions between the leader and his political party or the bureaucracy to explain how institutional changes could arise within the same party, albeit led by different leaders.
Interestingly, leaders could matter for growth beyond their tenure even if there is no significant difference in growth before and after the leader's tenure. The authors mentioned "the successful economic policies of Lee Kuan Yew in Singapore". Indeed, political leaders in Singapore have invested heavily and continuously in grooming subsequent generations of leaders as well as the bureaucracy to ensure that political instability would be minimized in the event of a leadership transition and that economic growth would be sustained. In this instance, the lack of any difference in growth rates before and after a leadership transition does not necessarily imply that leaders matter less for growth; here the impact of leaders simply works through a different avenue.
This comment has been removed by the author.
ReplyDeleteThis comment has been removed by the author.
ReplyDeleteAgain, not really an "Econ" person, but really enjoy the readings, Coase in particular. I suppose this is because I began with my standard eye rolling about the idea that that rancher or farmer could know all this information, think about it in those specific terms, and engage in a negotiation in the manner described. But then, BOOM, he brings in the cost of information, which maps really well onto my lived experience, where finding the right price and doing the appropriate inspections of a large purchase are, at a minimum, very time consuming and often monetarily costly as well.
ReplyDelete@ Kim and Man Yan: I learn a lot from your comments on Jones & Olken's article. After reading the article again by myself, I, too, am puzzled by the true mechanism between leaders and growth. While it is tempting to argue that the Chinese economy took off during the 1980s because Mao passed away, I will say China made it because the leadership chose the right policies, which was surely not an easy thing because the Chinese Communist Party suffered greatly internal struggle between the conservative faction (led by Chen Yun) and the reform faction (headed by Deng). Therefore, what we can imagine is if Deng failed to persuade the whole party leadership to adopt market economy, then the post-1980 Chinese history will be totally different. And Man Yan is surely right that the influence of a leader will continue even after he is gone, and there is still variation across leaders regarding their impact and control over policy-decision making that cannot be detected simply based on the difference between democracy and autocracy. So I guess two conclusions I have here are: One, it may be dangerous to give the leader too much credit. Two, democratic leaders may not necessarily weaker or less influential compared to their autocratic colleagues. I understand Jones and Olken did recognize these nuances, but I am still not satisfied with their results because I think there is simply more to say beyond statical results.
ReplyDeleteA few thoughts on the North and Weingast article and the ensuing debate:
ReplyDelete- OK, this is a minor side track, but the way in which rational choice is applied in trying to explain historical developments sometimes annoys me. Page 807: “In many of the simple repeated games studied in the literature, this incentive alone is sufficient to prevent reneging” – followed by some “rational” exceptions to this rule, in the example of the confiscation by Edward I of Jewish property in 1290. However, the explanation seems to me a post hoc rationalization. More immediate explanatory variables spring to mind: There was increasing anti-semitism in England since the resurgence of the blood libel in the 12th century, and also complaints about usury. Jews in England were the Sovereign’s personal property, and they could be taxed at will. An important reason why Jews mattered less economically at the time of confiscation in 1290 was that Edward I had already taxed them so heavily, so that there was not much left. Veight, whom North and Weingast cite, provides no evidence that Edward I in actuality considered the relative value of Jews against Italians; nor that he felt that the economic situation was such that he no longer needed them. In fact, Edward I constantly sought funds from any source he could find, and he used the expulsion of Jews, which was sought by Parliament, as an argument for getting more taxes from the English. There is every reason to believe that anti-semitism on the part of Parliament, probably Edward I (who, however, also seems to have felt some pressure from Parliament in this matter) played the major role of the decisions in this regard, leading to expulsion in 1304. Phillip II of France expelled Jews from crown lands, in 1192, John I of Brettany expelled them in 1239, and a little later Louis IX of France expelled them from France. All of them experienced revenue loss. It is possible to construct games where this course of action is rendered rational, but it transforms the concept of rationality into a truism without analytic or substantive meaning.
- I find North and Weingast’s general point to be relevant, and Stasavage’s reservations and elaborations to be generally useful. We know from many concrete situations that a parliament is not seen as a guarantee against expropriation. Indeed, in interwar Europe, the fair of radical, expropriationist parliaments abounded in Europe, and contributed to capital flight, class struggle and fascism. It is not unlikely that this was also the case in England and France in the 18th century.
However, it is difficult to say with any confidence from Stasavage’s account. The main evidence that he in my view presents, in addition to theoretical considerations, is the spike in interest rates during unified Tory rule from 1710 onwards. However, this period coincides with the last phase of the War of the Spanish Succesion. The costs of that war drove interest rates up everywhere. Unified Tory rule might have contributed, but we do not know. Whether London financiers and landowners behaved, or were believed to behave, differently especially in the English Parliament, Stasavage investigates to superficially. However, there is a huge debate about a similar question in the US Federal Convention. In general, it has been difficult to establish the effect of economic interests in this regard. There might be ideological or religious questions involved (repay your debt), but there are also other, more vague questions of interest involved. Bear in mind that we at any rate discuss a wealthy elite, though their sources of wealth differed. Arbitrariness was welcomed by neither financier nor landowner. The popular legislatures of the American colonies, on the other hand, with representation not only from the elites, but from the if not “the common man”, then at least people without personal wealth, tended to be less keen on repaying government debts.
- An example which lends contemporary relevance to the mechanism proposed by North and Weingast can be found on FrumForum today:
ReplyDelete"Some thoughts on the country he led, inspired by a recent visit:
Here’s a story I heard last week from a onetime foreign investor in Argentina. The thing that drove him out of the country was a 5 percent tax on his company. Five percent may not sound like much, but what mattered was not the amount of the tax. It was the way it was imposed. The tax was not enacted by Congress. It was not even ordered by the president.
One fine day, one of his operations received a letter from a government ministry with a sudden demand for payment. As far as he could tell, the demand was ungrounded in any law. He litigated the matter and (eventually) prevailed. But a country where the government could remake the rules at any time was no country for him."