Available now: USAvailable now: UK

Disrupting the Clientelistic Equilibrium in Chile  

Thus far we have examined various shocks to a clientelistic equilibrium that might disrupt it. The entry of new political forces which mobilize outside the system (Shefter), democracy (Lizzeri and Persico) and changed economic opportunities (Chubb). 

Another possibility is that changes in political institutions at a finer level than democracy could influence the incentives to be clientelistic. A large literature in political science argues that features of the electoral system can encourage clientelistic politics, because, for example, they make it easier to win office with a small number of votes, thus encouraging patronage rather than public goods provision.

A classic paper on this is by John Carey and Matthew Shugart, “Incentives to Cultivate a Personal Vote”. Carey and Shugart coded different electoral system on the extent to which they encouraged clientelistic strategies (Colombia won, or lost, depending on your perspective). The results suggest that, even ignoring such big transitions as movements towards democracy, changing things like the electoral system within a democracy might have large effects on clientelism. Though we are not aware of any empirical work that conclusively shows whether changing electoral rules does reduce clientelism, there is some research that is relevant here. First, as we have already discussed, Thomas Fujiwara’s research establishes how the change to an electronic voting system in Brazil made it easier for illiterates to vote, with a significant impact on the political equilibrium.

Second and more relevant, Jean-Marie Baland and James Robinson’s “Land and Power: Theory and Evidence from Chile” studied the introduction of an effective secret ballot in Chile in 1958. Before that, landowners, who could observe the way that their workers voted, coerced them into supporting right-wing political parties. This was particularly true of a class of workers called inquilinos who were resident on large farms and thus particularly subject to landlord control. When the secret ballot was finally introduced, the vote share of right wing political parties fell differentially in areas where inquilinos were an important part of the labor force (or electorate). Here the change in balloting procedures freed the inquilinos to vote for who they wanted, thus breaking down the clientelistic control of the labor force. Sometimes reforming the details of political institutions can have a major impact of patronage politics.  


Disrupting the Dysfunctional Equilibrium in Naples  

In our discussion of factors that might create a transition from patronage to programmatic politics, our focus has been on politics, the emergence of new political parties that form outside the system of patronage (Shefter) or the transition to democracy (Lizzeri and Persico). Obviously, it’s not just politics but also economics that shapes the feasibility of transitioning out of patronage politics.

As Avinash Dixit and John Londregan pointed out in “The determinants of success of special interests in redistributive politics”, the poor value patronage more than the rich. This is intuitive (but of course it does depend on assumptions). A natural assumption is a diminishing marginal utility of income and a constant disutility of switching one’s vote in response to monetary rewards (or a disutility that increases with income). This would then imply that when someone is poor, the offer to buy their vote or give a job creates a larger increase in utility than when someone is rich. To get the same increase in utility a rich person would have to be offered more money or a higher wage (and with a constant disutility of switching one’s vote, the poor and the rich have to be given the same increase in utility for purposes of patronage politics). All of this then leads to a simple conclusion: patronage will be more effective and cheaper when targeted to the poor.

One consequence of this reasoning is that the poor’s economic options might have a first-order impact on the cost and feasibility of patronage. If so, one way of ending patronage politics might be to improve the economic opportunities of the poor and of other target groups of patronage.

An interesting example of this is in political scientist Judith Chubb’s Patronage, Power, and Poverty in Southern Italy: A Tale of Two Cities. Chubb’s book is a wonderful analysis of how the patronage machine of the Italian Christian Democratic Party functioned. It provides powerful examples of how the party bought votes and how they made sure that people voted as they agreed. For instance, prior to an election a person would be given a left shoe and told how to vote. If they went through with the agreement after the election, then they got the right shoe. The genius of the whole thing was of course that, to both the voter and the party, one shoe was useless without the other, thus helping to cement the deal on both sides. 

Chubb’s analysis also explains how this system could function in a democracy even with a large number of voters. One reason was that though giving patronage employment was expensive, it didn’t just give the Christian Democrats one vote for each job, but the vote of the whole extended family connected to the employed person. By having a relative employed, these people got preferential access to many scarce resources, for example public healthcare. Such access didn’t cost anything (they just got ahead of other people in the queue), but it was very effective in ensuring political support.

Chubb’s book is split into two parts. The first shows this machine at work at full effectiveness in Palermo. The second part, more relevant to this post, shows how it crumbled in Naples in the 1970s. After the discussion of Palermo, it is quite a shock that this complex and very effective patronage machinery, which seemed to have every angle covered, just collapsed. How did it happen?

What Chubb shows is how important simple economic opportunities can be. In Naples, a regional development policy emanating from Rome led to the construction of car factories. These created new jobs which were relatively well paid, reducing dependence on patronage. Possibly also important, the car factories were taken over by northern communist trade unions. The combination of high paid jobs and unions was a double whammy for the Christian Democrats. The unions ended their control of appointments in the car factories while the higher wages meant that people were not so ready to sell their vote (or at least according to Dixit and Londregan’s logic which we just discussed) would have demanded a higher price, making this strategy more expensive and less feasible for the Christian Democrats). The result was a significant loosening of their grip on local politics, and in 1975 they fell from power.


Democracy and the Transition to Programmatic Politics  

After a long detour, we return to our discussion of the transition away from patronage politics.

Martin Shefter’s thesis about the transition to programmatic politics, which we discussed here and here, is about how patronage politics may be undermined by the entry of new political parties in a democracy. But shouldn’t democracy, and more generally the nature of political institutions, influence the incentive to use patronage politics? Perhaps democratization would change the incentives or ability of politicians to use patronage?

There are several ideas in the literature about whether and how democracy may reduce patronage. One of these follows from our earlier discussion here in the context of Robert Bates’s seminal book Markets and States in Tropical Africa. Politicians often prefer to curry favor through patronage, which can be targeted, to providing public goods, which cannot. If democratic institutions restrict the extent to which such targeting is possible, they will naturally tilt the balance towards providing public goods in order to win support.

Another specific mechanism has been developed in the interesting paper by Alessandro Lizzeri and Nicola Persico “Why Did the Elites Extend the Suffrage?”. They argue that the trade-off between public goods and patronage depends on political institutions and in particular on the extent of the voting franchise. If voting rights are narrowly distributed, for example as they were in Britain prior to the 1867 Second Reform Act, then in order to win a politician has to only attract relatively few people’s support. Patronage, which can be targeted specifically to these people, is then both feasible and effective. When voting rights are expanded, however, it gets more and more costly to provide patronage to win elections. Instead, the option of providing public goods more broadly in order to appease a larger segment of the population becomes increasingly attractive.

Lizzeri and Persico not only show that democracy will be associated with a reduction in patronage but also that, through this effect, non-democratic elites may in fact prefer to create democracy to maintain a narrow voting franchise. This is because, with a narrow franchise, it is politically rational for a politician to rely on patronage to win elections. But the same is true for competing politicians, who can promise patronage. The resulting competition would then resemble a prisoner’s dilemma — politicians outcompeting each other in promising patronage to a narrow group. Think of it this way: if a politician tries to win an election by offering public goods, another one can do better by offering patronage, which is more effective and cheaper. Because such prisoner’s dilemmas are costly to the participants, the elite politicians themselves may be better off when the voting franchise is extended and the patronage option is off the table. (Lizzeri and Persico’s is in turn an alternative but complementary theory to ours, developed in the paper “Why Did the West Extend the Franchise?” and in greater detail in our book Economic Origins of Dictatorship and Democracy; we’ll return to these ideas soon).

Of course the creation of democracy does not necessarily remove patronage politics. Political machines can adapt. But Lizzeri and Persico provide a solid reason to believe that there is one important mechanism which pushes democracy away from patronage politics.


Response to Jeffrey Sachs

Several people asked us why we haven’t responded to Jeffrey Sachs’s review of Why Nations Fail in Foreign Affairs. Well the answer was sort of in-between the lines in our response to Arvind Subramanian review (the original review is here and our response is here): we said that thoughtful reviews deserve thoughtful answers. What about not-so-thoughtful ones?

Be that as it may. We cave in to pressure.

Sachs charges that we are “simplistic” and our argument “contains a number of conceptual shortcomings”. But in each case, these are either just stated (and are wrong) or he is criticizing something we haven’t said. The Sachs strategy seems to be to throw a lot of mud, hoping that some of it would stick — did we say that we didn’t think it was quite thoughtful?

Let’s go through each one of his points in turn.

1. Sachs says: “dictators have sometimes acted as agents of deep economic reforms, often because international threats forced their hands.”

Perhaps we are not as deferential to dictators as Sachs would like us to be, but this is very much what we argue in our discussion of growth under extractive institutions. Such growth takes place when elites find it in their interest to allow new technologies and institutional changes necessary for economic growth. The entire Chapter 5 is devoted to this, and we return back to this issue several times in the book, including the last chapter.

2. He continues: “The authors also conflate the incentives for technological innovation and those for technological diffusion.”

We do no such thing. We emphasize that growth under extractive institutions is especially feasible, as in China today, when it can proceed rapidly by importing existing technologies from other economies. One of our central arguments is that inclusive institutions are necessary for sustained innovation, but import of technology can sometimes take place under extractive institutions. Does this look like ignoring the difference between innovation and diffusion?

We also go to pains to discuss how, when they feel threatened, rulers and elites in Ming and Qing China, the Ottoman Empire, and 19th-century Russia and Austria-Hungary have opposed the diffusion of technologies. The point we make is that innovation does require inclusive institutions but extractive institutions, though they sometimes allow the use of existing technologies, will often also block the import of technology because this too can be threatening to existing power-holders.

3. He goes on: “What’s more, authoritarian political institutions, such as China’s, can sometimes speed, rather than impede, technological inflows.”

This is a fair point, which one of us has argued theoretically and empirically in past work, but at the end, whether catch-up growth under extractive political institutions can be as fast or actually a little faster than growth under inclusive political institutions is secondary for anything we discuss in Why Nations Fail.

4. Sachs then charges: “The book misinterprets the causes of growth in another way…. a state’s power depends… on adequate resource base…”

Well, not really. There is no evidence we are aware of that a state’s powers depend on resources. Sure, Sachs himself has run some kitchen sink growth regressions where some geography-related variables are significant correlates of growth (but state’s power? We have never heard him to make that point before). In any case, these regressions do not really stand up to scrutiny— and of course are notoriously ill identified to say the least.

5. And then: “Not only can unfavorable geography cripple states; it can also slow the development and diffusion of technology.”

Again, evidence… (and see our elaboration of this in the next point).

6. Then comes the coup de grace: “The overreaching effect of these analytic shortcomings is that when Acemoglu and Robinson purport to explain why nations fail to grow, they act like doctors trying to confront many different illnesses with only one diagnosis.”

Well, we are not doctors. Sachs probably thinks he is one (though we didn’t think his doctorate was a medical one). Our purpose was not to write a doctor’s or even a practitioner’s manual, but provide a framework. We think, and perhaps Sachs disagrees, a framework that says there are 17 factors, each of them hugely important is no framework at all. The power of a framework comes from its ability to focus on the most important elements at the exclusion of the rest and in so doing in providing a way of thinking about these elements, how they function, how they have come about, and how they change. For us those elements were related to institutions and politics, and we have focused on them.

7. Sachs then charges that “Extraordinary claims require extraordinary evidence. Yet Acemoglu and Robinson do nothing of the sort” and argues that we provide no evidence.

Right, we do not in the book. But that’s because a book for a general audience is not the right forum for presenting academic research, and we spent many years of our lives precisely on writing academic papers providing exactly the sort of evidence.

Funny that Sachs ignores this. One of our papers is even called “Reversal of Fortune: Geography and Institutions in the Making of the Modern World Income Distribution”, which sounds as if it might have something to do with the competing geography and institutions hypotheses. What’s more, another one of those papers, “The Colonial Origins of Comparative Development” shows precisely the type of historical persistence in colonial institutions that is the basis of our framework in Why Nations Fail, and shows that, at least in the sample of former European colonies, once the effect of institutions is properly controlled for, there is no room for geography in explaining the huge cross-country differences in income per capita.

Yes, it is funny that Jeffrey Sachs ignores this because when this paper came out, he was so troubled by it that he decided to write a comment on it. He asked us for our data and rushed to produce what he thought was a devastating comment (which, interestingly, did not really question our approach).

It was also funny (or sad) that Sachs couldn’t even bother to do the regression analysis with the data we gave him properly. Our argument was about institutional variation within colonies but he included Britain and France in his sample, the colonizers, obtaining nonsensical results! (See the discussion in here). Funny that we haven’t heard about that paper since. And funny that Sachs himself seems to have forgotten about it too.

And these are not the only relevant pieces of research. Other works by us and others have also shown that once the historical role of institutions is properly factored in, geography doesn’t seem to matter at all or much. So yes, we don’t provide the econometric evidence in the book, which isn’t of course the right place to do it, but econometric evidence is abundantly loud in the way it speaks on these topics.

8. Sachs goes on: “There are also countries that possess both inclusive political and inclusive economic institutions yet never achieved much development, often due to geographic barriers” and mentions Botswana, and claims that it was fated to stagnation at independence, and its growth is just due to diamonds, which we (supposedly) ignore.

Well, again, we do nothing of the sort. In Why Nations Fail, we write:

When the diamonds came on stream in the 1970s, they did not lead to civil war, but provided a strong fiscal base for the government, which would use the revenues to invest in public services. There was much less incentive to challenge or overthrow the government and control the state. Inclusive political institutions bred political stability and supported inclusive economic institutions. (Page 413)

It sure doesn’t look like we don’t know Botswana has diamonds, does it? In fact, we have written a paper entitled “An African Success Story: Botswana” on Botswana’s growth experience, much of it about diamonds. And of course, diamonds are important for Botswana’s growth as we note in that paper and in Why Nations Fail, but Sierra Leone and Angola also have diamonds, and Nigeria has plenty of oil.

What’s funny here again is that Jeffrey Sachs himself has argued that natural resources are a curse and a major contributor to the disappointing growth performance of countries like Nigeria, Sierra Leone and Angola, and now he says diamonds are a blessing and Botswana is rich because of diamonds? Odd…

9. Sachs then says we have Nogales all wrong, because rather than showing the futility of geographic explanations, which Sachs loves of course, Nogales is all about geography. He says “the case of the two Nogaleses is about geography and nothing else. Only geography can explain why the desert city of Nogales, Sonora, even exists”.

Well, we don’t know where to start. First, as we explain in the book, Nogales exists because it formed on the border between the US and Mexico which was defined by the Gadsden purchase of 1853. By 1880 there was a trading post on the border at the location of present day Nogales and by 1883 a US post office. So Nogales exists for historical reasons related very much to a political boundary.

Second, we never said that geography is irrelevant — how would one otherwise explain why there aren’t holiday resorts in Antarctica? We argued and demonstrated that the geography hypothesis, which links the huge cross-country differences in prosperity to geography is wrong and unhelpful — precisely the geography hypothesis that Jeffrey Sachs himself endorsed and argued, for example here.

So what Sachs is doing here is to shift the goalposts by claiming that Nogales proves geography matters. If by “geography matters” he means that a border town next to a rich country is more likely to develop than other parts of the country, that’s an entirely different proposition from the geography hypothesis which claims that geographical factors are a crucial determinant of cross-country or cross-region income differences. The latter would have no explanation for Nogales Arizona being so much richer than Nogales Sonora, and the latter is the one that Jeffrey Sachs used to espouse, but perhaps no longer.

Of course geography can explain that for those planning to open a business in Mexico that will import technology or trade with the United States, border towns such as Nogales Sonora are quite attractive. But that really deepens the challenge to the geography hypothesis. Nogales Sonora is poor despite that advantage, a point we make quite clearly in the book, but Sachs conveniently chooses to ignore.

10. Sachs continues: “Yet Acemoglu and Robinson seem generally unwilling to think dynamically in spatial terms.”

We don’t quite know what “think dynamically in spatial terms” means, but while Sachs was talking about some places being permanently disadvantaged, which doesn’t seem very dynamic, we distinguished in “Reversal of Fortune: Geography and Institutions in the Making of the Modern World Income Distribution” between the simple and sophisticated geography hypotheses, the latter allowing the importance of geography to depend on time and the state of technology, and provided evidence against it being a crucial determinant of the historical evolution of prosperity differences across countries. Moreover, our book approaches world inequality by looking precisely at the historical dynamics of institutions and explains, for instance, the poverty of Africa by the fact that it has been caught in an institutional vicious circle factoring in historical processes of state formation, the slave trade and colonialism. To us this seems more dynamic — and incidentally also hopeful — than trying to establish that Africa is condemned to poverty by latitude and malaria.

11. Sachs also asserts: “Outside Europe, in the 19th century, industrialization spread will successfully to places with good geography”.

Really? Such as Argentina and Uruguay? What’s so great about Australia’s geography, largely cut off from the rest of the world, with vast swathes of desert and tropics, very sparsely settled with huge transportation challenges? Indeed a famous historical account of Australia is called “The Tyranny of Distance”. What’s so much better about Mauritius’s geography, the biggest success with respect to industrialization in Africa? And what is it that is so bad about Mexico’s geography? Whatever that was, it did not prevent Mexico from building one of the most complex civilizations before Columbus as we argue in Why Nations Fail and have argued elsewhere.

12. Sachs then charges: “As for the future of development, Acemoglu and Robinson’s narrow focus on political institutions offers insufficient predictive help…. At the start of 1980, an economist basing his judgment on future economic performance on political and civil rights during the preceding decade or so might have foolishly bet on Gambia, Ecuador, or Suriname and almost entirely missed the rapid growth of authoritarian used Asia, most notably China.”

We much rather leave the predictive game to Sachs who, for example, seemed to think that IMF style adjustment policies (combined with repression of any opposition) he advocated in Bolivia were the secret to growth. Actually, perhaps he is right in some twisted way. After the reforms Sachs advocated got all of them fired, Bolivian tin miners went off to the Chapare Valley in Eastern Bolivia to grow coca. That’s what helped provide the social basis of the coca growers union which then became a key political base for now president Evo Morales and his MAS party. As we argued in a previous post that there might be some hope that the MAS party is changing Bolivian institutions in an inclusive direction. So perhaps Sachs has played an important role in this, but of course not through the mechanism he had in mind.

But we digress. Actually, Sachs is again misrepresenting our views. First, much of the book is about the difficulty of building inclusive institutions out of the ashes of extractive ones. So Gambia, Ecuador and Suriname would not have been our ideal inclusive societies, perhaps they are Sachs’s. Moreover, we put a lot of emphasis throughout, as we have already noted, on extractive growth in Chapters 14 and 15.

When we talk about future growth, in contrast to Sachs we first emphasize the difficulty of making any sort of prediction. We then point out that rapid growth is likely to come from several countries such as Ethiopia and Rwanda having achieved some degree of political centralization now jumping on the extractive growth bandwagon — not from Gambia, Ecuador and Suriname.

Did we say that we did not think Sachs’s review was thoughtful?


Are We Getting Sad and Old?  

Obviously the simple answer to this question is yes.

The bigger issue however is whether our defense of the House of Lords in our post here, albeit somewhat strained and understated, means that age has overcome the radicalism of youth and we’ve bought into and been co-opted by the absurd trappings of the British institutionalist status quo. Perhaps Jim is working on his knighthood? (Despite having spent seven years in the UK, Daron wouldn’t qualify for one).

This is certainly a possibility. But (thankfully) there is another. Though it is easy to criticize British (and other) institutions piecemeal, social scientists — and certainly we — don’t yet understand how systems of institutions fit together. It is tempting to argue that this institution is bad, that institution is good, and on the basis of this, one can even advocate changes and reforms.

But we don’t really know how the big picture fits together. So there is a danger in all sorts of reforms.

In defense of our radical credentials we should point out that we showed in joint work with Davide Cantoni and Simon Johnson, “The Consequences of Radical Reform: The French Revolution”, that parts of Europe which were reformed by French armies after the French Revolution grew faster in the 19th century than parts that were not so reformed. In this instance, radical reform worked and that however complex their motivations, French armies moved institutions in places they conquered in a more inclusive direction (though in the process they did also create a lot of upheaval and carnage to be included in the cost column).

The concern that reform — especially radical reform — of a complex set of interacting and dynamically changing institutions might backfire could lead one to sympathize with Edmund Burke who viewed the French Revolution with great trepidation. In his Reflections on the Revolution in France, Burke argued that (page 152):

It is with infinite caution that any man should venture upon pulling down an edifice, which has answered in any tolerable degree for ages the common purposes of society, or on building it up again without having models and patterns of approved utility before his eyes” Burke (1969)

(But before one gets too carried away with the virtues of Burkean conservatism, it is also worth remembering that Burke was generally opposed to democracy and did not have any such trepidation in taking radical action against France or Ireland).

In any case, a plausible argument can be made that British institutions have managed — despite their many imperfections that one is tempted to point out and despite their total compatibility with repressive and extractive colonialism when this was in the interests of the British elite – to create one of the world’s most durable inclusive and economically successful societies. Compared to places like Colombia, Haiti or the Democratic Republic of the Congo, perhaps it is true that the nexus of British institutions have some flexibility and a resilient core that has facilitated the gradual buildup of inclusive institutions. Perhaps it is also the case that ad hoc reforms, even when motivated by well-intentioned fervor against extractive institutions, may have unintended consequences because they will often start changing parts of institutions that have somehow come to fit together and interrelate in ways that neither we nor social scientists more broadly nor the reformers and the revolutionaries understand. OK call us conservatives!