At the American Interest, Arvind Subramanian reviews Daron Acemoglu’s and James Robinson’s Why Nations Fail. It’s a very thoughtful and engaging review; it veers away from the usual approach that other reviewers have taken, which usually pits one theory of wealth bifurcation against another, and instead comments on institutional theory itself. Subramanian’s critique is summed up in one paragraph, excerpted below,

The choice of axes is very important because WNF asserts that causation runs from politics (the independent variable on the x-axis) to economic development (the dependent variable on the y-axis). The authors are unsympathetic to causation running the other way. That is, they reject the modernization hypothesis, which asserts that improvements in standards of living will lead to more democratic politics, stemming for example, from increased demand for political freedom and participation. For Acemoglu and Robinson, political institutions bear the deep imprints of history, and although they are not immutable, their susceptibility to change induced by economic development is limited.

I sympathize heavily with Subramanian’s argument. There are times when Acemoglu and Robinson (A&R) don’t consider the implications of reverse causality, where “economics” influences political institutions. My own research on ~1990–2008 Somalia works in a similar direction, although it is a bit more radical (societies can create various types of justice-distributing institutions). On the other hand, I didn’t get the impression that A&R are “unsympathetic” to a reverse causality (in fact, in a paper co-authored with Simon Johnson, “Institutions as the Fundamental Cause of Long-Run Growth,” they do, in a sense, explicitly accept causality running from economics to politics when they consider the distribution of resources as a relevant determinant of political institutions).

The last sentence of the above-quoted excerpt doesn’t fit many of the examples A&R provide in the book. For instance, when they discuss England (and Great Britain more generally) there is a certain degree of causality running in the direction Subramanian prefers. The authors bring up merchant class pressure on parliament and the monarchy to change laws on how innovation is introduced to the market, whereas prior to this political change much innovation was stifled by the crown. It’s true that, in other ways, A&R paint it as a political transition (monarchy_t → revolution → monarchy_t+1), but they also argue that these events are historically contingent.

A&R do write that present institutions find their roots in a deep past, but I didn’t interpret this as a claim that political institutions’ susceptibility to change is limited. In fact, this flies in the face of a lot of the examples given throughout the book. Rather, I understand A&R’s political theory as one of an evolutionary process. The shape of modern governments is, partially, determined by what it looked like in the past; that is, we can only understand the nature of modern States by looking at what they evolved from. We can have an example of an extremely volatile government, and A&R’s statement would remain just as true.

Subramanian illustrates his thesis with the following graph,

Source: Arvind Subramanian, “Which Nations Failed.”

The countries shown on the graph, China and India, are the key examples to Subramanian’s argument.

Unfortunately, Subramanian doesn’t engage A&R’s discussion of China in Why Nations Fail. If I remember correctly, while A&R do mention China’s gradual, inelastic embrace of liberalization, they also try to explain the country’s meteoric GDP growth through two principal causal explanations,

1. Extractive institutions can still see GDP growth, but this is limited;
2. Developing economies have the “advantage” of being able to import technologies from advanced economies, making the process of catching up much quicker.

Whether China is a case against A&R’s theory is something that we’ll have to see in the future. But, the scope for their thesis to be proven wrong by China is fairly narrow. China would essentially need to lead the global economy, and China’s institutions will have to have remained very extractive.

India is a more powerful counter-example, and the take-away should be: democracy is not a panacea. Furthermore, one could make the stronger claim that there is no guarantee that a democracy will be predominately inclusive (I, for the most part, even reject the inclusive ↔ extractive spectrum, preferring the no extraction [anarchy] ↔ socialism [public ownership of the means of production] spectrum). We are still in an era where democracy is usually associated with certain economic and political liberties, and the truth is that the process of the universalization of these liberties goes hand-in-hand with the development of democracy. There are certainly governments that are politically more inclusive than others (e.g. democracy vs. absolute monarchy), but political inclusivity doesn’t always mean economic inclusivity.

T.H. Marshall, in “Citizenship and Social Class,” interprets the evolution of rights in this way: civil rights → political rights → social rights. Marshall’s article is both the seminal paper on the topic and highly controversial, but it gives some good insight on how we should interpret democracy. Democratic institutions are usually associated with the process of spreading political rights to a growing number of citizenry. For the sake of simplicity — I’m not sure Marshall would entirely agree with the representation here —, we can redress civil and social as negative and positive rights, respectively. As such, most the history of democratic institutions deals with the spread of negative rights; more accurately, I’d say that the main kind of right society has fought for are those which emancipate the individual from the State.

Since the dawn of the 20^th century, for the most part, social rights have become relevant. Again, while this doesn’t follow Marshall precisely, I’d look at social rights as those which guarantee some kind of entitlement. In modern liberal political theory (Rawls, et cetera), the purpose of these entitlements is to provide a certain equality in the political playing field, avoiding disparities of power in political discourse. But, these rights, by their very nature, are extractive: they require the redistribution of wealth. While liberal political theory tends to be exceptionally strict about the extent of social rights, real world politics don’t often work this way (and political theory can be contingent on the status of real world politics — political theorists are influenced by the prevailing zeitgeist). There is a historical trend of an ever-growing welfare state, at the expense of the market economy.

I’m not commenting on the justifications for these policies, rather simply pointing out that there is a (nonlinear and noncontinuous) negative relationship between entitlements and economic growth. There is a point where, even in terms of A&R’s inclusive ↔ extractive spectrum, democracy can become more extractive than inclusive. It’s important to temper the strict equivocation between extreme welfare-states and extractive governments by considering the fact that politics isn’t necessarily homogeneous. Policies are heterogeneous, and some are exclusive and others are inclusive. Interpreting this within the theme of welfare-state democracies, there is mutually incongruent legislation which will impact economic growth in different ways. The most obvious example are Nordic states, such as Sweden, which are well-known for their copious (and, I don’t mean to use this word in a non-neutral way) welfare programs, but also enjoy a relatively strong private sector.

Nevertheless, India is a strong example of a society enjoying a polity generally described as inclusive, yet suffering from relatively weak economic growth. At times, India has been taken as an example of a strong developing economy, but it teeters on that edge between success and failure. India is also a strong example of why ideas do matter; A&R are hostile towards the notion that ideology impacts economic development. A&R can respond by pointing out that India’s government isn’t corruption-free, and that democracy in name doesn’t translate to democracy in practice (although, Indian democracy is no doubt closer to its essence than, say, Venezuelan democracy under Hugo Chavez), and there certainly is a substantial element of truth in the claim. But, democracies still operate through political force, and a majority ideology doesn’t necessarily correspond with what correlates with strong economic growth.

There is evidence that Subramanian missed out on a lot of A&R’s arguments in Why Nations Fail. He ascribes to them a stronger stance on the direction of causality between political institutions and the market than they actually take. But, the review does make several good points, and it ought to lead us to other, stronger re-interpretations of A&R’s work. More than criticism, I find Subramanian’s review a good reminder of the fact that there’s a lot more to institutions and their relationship with economics than what has been uncovered so far.

[Edit: Acemoglu and Robinson respond to Subramanian's review: "China, India, and All That."]