Making sense of the 2024 Economics Nobel

The 2024 Nobel Prize in Economics has been awarded to Acemoglu, Johnson and Robinson “for studies of how institutions are formed and affect prosperity”. In this post, Pulapre Balakrishnan evaluates the econometric approach and representation of colonialism – as linked to institutions – in some of their prominent works. Balakrishnan contends that any account of economic phenomena must commence with a theoretical model including a clearly specified motive force; in his view, this is lacking in the laureates’ contribution.

“In spite of the vogue of ‘Quantitative Economic History’, economic historians are under less temptation to see their subject as purely quantitative.”
– John Hicks (1969)

The 2024 Sveriges Riksbank Prize in Economic Sciences has been conferred jointly on Daron Acemoglu, Simon Johnson and James Robinson (henceforth, AJR). They have co-authored several articles and books in various combinations that include one or the other among them, but have together written two well-known articles. The first is “The Colonial Origins of Comparative Development” (2001) and the other one is “Institutions as the Fundamental Cause of Growth” (2005). My assessment of their contribution is based on these two studies.

In AJR (2005), where the authors reflect on their earlier work together, they state that their objective is to answer the question of why some countries grow faster than others. While conventional growth theory focuses on the proximate determinants of growth – think Solow – they themselves wish to go back a step and address the question “what causes growth?” They consider earlier contributions as being concerned with the “mechanics of growth” – think Lucas here – while they themselves are unravelling something deeper. To convey more clearly what they attempt, they refer to Douglass North’s observation that capital formation, economies of scale, etc. are “already growth”. It may be said that AJR are aiming at a ‘meta narrative’. ¹

The institutions and growth hypothesis

Methodologically, theirs is a model-free investigation, attempting a parsimonious explanation of growth globally, that is, across countries. Their starting point is to postulate that whether or not an economy grows depends upon how it is organised, that is, on its economic institutions. Institutions themselves are the rules governing interaction in society. As they proceed, we find that AJR have in mind not only interaction among citizens but also between State and citizen. So, institutions are “the rules of the game”, as described by Douglass North. AJR’s main contention is that “There is convincing empirical support for the hypothesis that differences in economic institutions, rather than geography or culture cause differences in incomes per capita” (AJR 2005, p. 402; italics as in the original).

They arrive at this conclusion by exploiting what they see as institutional divergence between the countries colonised by Europeans in the ‘modern age’. Briefly, Europeans settled in the sparsely populated regions, putting in place “good” institutions. In the densely populated ones, they imposed “extractive” institutions. They state “… we think of good economic institutions as those that provide security of property rights ….” (AJR 2005, p. 395) They do not define extractive institutions as precisely, though their econometric exercise conveys what they have in mind, which is the absence of “protection from expropriation”. Anyhow, for AJR, these colonial institutions have remained till today, and explain the current divergence in income among the ex-colonies – those with the good institutions having achieved far higher growth than the others.

Assessing the econometric approach

That secure property rights, assuring that the returns to effort and/or investment will not be expropriated, are a necessary condition for an economy to achieve long-term growth is not something that any sensible economist would contest. However, note that AJR claim that secure property rights “cause” growth. Their claim is based on their implementation of an ‘instrumental variable’ (IV)² regression that instruments institutional quality with European settler mortality at the time of colonisation. Without stating it in as many words, AJR view their empirical strategy as mimicking a randomised controlled trial (RCT)³ and their method approximating the standards of explanation in the natural sciences. In the rest of this post, I shall first comment on their econometric results and then evaluate their representation of colonialism.

The scientific community (in the sense of Thomas Kuhn) to which AJR belong, view their choice of instrument as an inspired one. It satisfies the conditions of both ‘externality’ and ‘exogeneity’ (Deaton 2010). There can be no question of reverse causality here, as current national income cannot cause settler mortality rates of several centuries ago. Prima facie, the econometric estimate appears robust but is less so upon closer scrutiny. For the first-stage regression of institutional quality today on settler mortality, AJR (2005) assert that there is a “…. very strong relationship between the historical mortality risk faced by Europeans and the current extent to which property rights are enforced. A bivariate regression has an R-squared of .26” (AJR 2005, p. 417). But an R-squared of 0.26 does not come across as particularly strong even for a cross-section regression.

A second point relates to the empirical strategy AJR adopt to test their hypothesis. Their econometric specification relates country per capita national income in 1995 to the average degree of protection against the risk of expropriation over 1985-1995. For them the latter is the main variable of interest in the regression, which also contains controls. Their prior is that the relationship between these two variables is positive, and that is what they find. However, a scatter plot between the two variables, given in Figure 1 of AJR (2005), reveals that the relationship is not monotonic. For instance, there are only 18 countries in a sample of about 100 that have better property rights protection than India does, but 73 have a higher income. In addition, visual inspection of the same scatter plot shows high dispersal, with more countries further away from the line of best fit than closer to it. This suggests that institutional quality can be taken as only one factor in economic development. Overall, the behavioural foundations of the AJR hypothesis are not developed by its authors, leaving their econometric specification ad hoc. This would be a consideration when evaluating their econometric results.

Third, even if the econometric grounding of their study were much stronger, it may still be asked what causal mechanism links growth to property rights. In RCTs in medicine and the biological sciences more generally, it may be that the outcome of an experiment is viewed as having been caused by the treatment, but this cannot be deemed sufficient in the behavioural sciences, of which economics is a part. A causal mechanism needs to be specified. While claiming to explain comparative development, AJR have no model of development. It is interesting to find that they did not feel the need to provide one, perhaps having assumed that the very prestige of the method of casual inference precludes the need for it.

Representation of colonialism

To bring up the issue of the causal mechanism in the AJR account is a matter of more than just academic interest. Their reducing colonialism to a matter of institutions serves to mask other possible causes. They speak of “understanding the colonial experiment” (AJR 2005), but it may be asked if they have one. Arguably, the institutions were only incidental to the objective of the coloniser. Slavery would be a case in point. Its existence in the United States even after the American republic is particularly problematic for the AJR thesis, for it is a case of secure property rights (of the slave-owner, that is) coexisting with extraction. Now, their neat classification scheme distinguishing countries on the basis of good and extractive institutions breaks down. The case of slavery points to the possibility that institutions came second under colonialism.

More importantly, institutions were not the only means by which the colonisers achieved their objective. In some important instances, they achieved their objective by directing income flows. Institutions may have little to do with this aspect of colonialism, perhaps its most defining one. An astute account of the role of income flows under colonialism is to be found in the writings of Keynes (1930): “…. I trace the beginnings of British foreign investment to the treasure which Drake stole from Spain in 1580. In that year he returned to England bringing with him the prodigious spoils of the Golden Hind. Queen Elizabeth was a considerable shareholder in the syndicate which had financed the expedition. Out of her share she paid off the whole of England’s foreign debt, balanced her Budget, and found herself with about £40,000 in hand. This she invested in the Levant Company, which prospered. Out of the profits of the Levant Company, the East India Company was founded; and the profits of this great enterprise were the foundation of England’s subsequent foreign investment.”

The case of India

A focus on institutions without taking income flows into account is to misunderstand what AJR refer to as “the colonial experiment” and leaves one unable to make a quantitative assessment of what was at stake for the European colonisers. This can be illustrated by looking at the relationship between India and England. Unlike in AJR, the colonial experience was not uniform, and to view it through a single frame, such as the institutions, would be procrustean. Now studying the Indian case makes sense as it constituted a very large part of the whole European colonial experience. It is this treatment of the European colonies as alike in all respects except their institutions that leads AJR to suggest that the colonies of North America and Australia-New Zealand were successful in industrialising while India and Bangladesh were not. This is blind to the feature that for the most successful European coloniser, namely Britain, its white and coloured colonies were not at par. The former were seen as extensions of the mother country, to be nurtured, while the latter were possessions acquired purely for their exploitation. In this account, Britain industrialised at India’s expense by first extracting surplus from India and then by using it as a defenceless market for its goods – ‘defenceless’ in that Britain itself had protected its markets from Indian manufacturers at the nascent stage of its industrial development while India could not have protected its industry. Irfan Habib has estimated that during a crucial phase of the Industrial Revolution outflows from India amounted to 30% of capital formation in Britain. This by no means small investible surplus was lost to India forever. Later, British textile exports to India are estimated to have amounted to two-thirds of domestic consumption in India (see Habib (1977) for the source). AJR claim to explain the ‘colonial origins’ of comparative development. What they do, however, is to compare development across the European colonies while ignoring the role of colonialism in the rise of Europe.

Finally, the rise of India after 1947 has implications for the AJR thesis. It would be highly plausible to argue that India broke out of colonial stagnation due to State direction. The Nehru-Mahalanobis Strategy shifted the economy to a higher level of output, from which resulted continuous and accelerating growth enabled by an evolving policy regime, most recently in the form of the economic reforms of 1991 (a causal explanation of this is provided by Balakrishnan, Das and Parameswaran (2017)). This progress of India’s economy was achieved at a time when there was no major overhaul of its economic institutions, though the forced extraction of surplus from its economy ended. It has been pointed out that China’s rise with ambiguous property rights is a challenge to AJR, but actually a more direct challenge is posed by India’s growth after 1947. India was intrinsic to European colonisation in a way that China was not. To cut a long story short, Britain’s rise and India’s falling back in the colonial era have little to do with institutions. It has to do with colonialism understood as unrequited income transfers from the colony to the metropolis. Under colonialism the extraction was enabled through military power. Institutions were secondary to this enterprise. Actually, the East India Company instituted property rights, which are foremost among AJR’s good institutions, precisely so that land revenue could be extracted efficiently for responsibility for it could be pinned down.

Conclusion

Shorn of the deft handling of the putative issue of endogeneity in their econometrics, there is little of interest to the wider economics profession in the work of Acemoglu, Johnson and Robinson. There is neither economic analysis nor historical narrative relevant to their field of investigation, namely, the colonial origins of comparative economic development. Above all, they provide no mechanism to back their claim that they have delivered a causal account of the role of institutions in economic growth. It would be reasonable to insist that any account of economic phenomena must commence with a theoretical model, not necessarily mathematical, which includes a clearly specified motive force.

Versions of this paper were presented at seminars at the Madras Institute of Development Studies, Chennai and Krea University, Sri City. I thank without implication Rohit James, M. Parameswaran, S. Sivakumar, M. Suresh Babu and Rishi Vyas for advice.

The views expressed in this post are solely those of the author, and do not necessarily reflect those of their organisations or of the I4I Editorial Board.

Notes:

Solow's growth theory posits that that long-term economic growth depends mainly on technological progress, as growth of capital and labour face diminishing returns.Lucas's theory of growth emphasises the role of human capital as the key driver of economic growth.
The instrumental variable strategy is used in empirical analysis to address concerns about endogeneity. The instrument is selected such that it correlates with the explanatory variable, but does not directly affect the outcome of interest. It can therefore be used to measure the causal relationship between the explanatory variable and the outcome of interest.
A randomised controlled trial is a method which aims to evaluate an intervention by randomly allocating participants into 'treatment' (receives the intervention) and 'control' (does not receive the intervention) groups.