Trevor Chow Blog Archive Newsletter About The Blog

Back-of-the-envelope low-quality musings: #takes


Longform stuff I've put time and thought into is at #effortposts.


To hear from academics and policymakers, #interviews.


Compiled scribbles on what I read can be found in #notes.


Topic-related tags include #economics, #coronavirus and #miscellaneous.

A Primer on Economic Growth: Part II - Electric Boogaloo

24 Jul 2020

In Part I, we saw what growth can be modeled as - but we were left wondering where that comes from. In Part II, we will be examining several hypotheses on the fundamental drivers of economic growth - that is, how the factor accumulation and productivity gains arise in the first place.

What is the theory of geographical determinism?

One theory pushed forward by Jared Diamond in Guns, Germs and Steel is geography. He argues that agriculture arose in Mesopotamia, Mesoamerica and China initially due to the relative scarcity of game, nuts and berries, while the soil was fertile, motivating new forms of food production. This resulted in humans trying to plant seeds and domesticate wild animals, using them to assist with agricultural work. Agriculture spread across the world, with archaeological data showing that agricultural innovations diffusing east and west far sooner than north and south, due to areas of similar latitude sharing a similar climate and environment.

These agricultural societies developed immunities to deadly diseases, because of the constant proximity to domesticated animals and increased population density. The structure of agriculture society put a premium on a writing system, causing written language to emerge and spread. The ability to store crops created leisure time, and the increased population density allowed the faster spread of ideas - these combined to result in faster technological growth. Agricultural societies tended to develop or combine into larger groupings, which required the centralisation of power, resulting in the rise of political structures and states.

Combined together, basic geographical differences between early societies magnified over time, leading to vast differences between societies’ health, technology, and social structure. These in turn created tremendous advantages that affected how quickly they grew and conquered. When applied to the reasons for Eurasian growth ahead of American, there are four factors. Firstly, the continental differences in the plants and animals available for domestication. Secondly, the rate of diffusion due to the geographic orientation of Europe and Asia (east-west) compared to the Americas (north-south). Thirdly, the ease of intercontinental diffusion between Europe, Asia, and Africa. Fourthly, the differences in continental size, which led to differences in total population size and technology diffusion.

Doesn’t this mean there’s no possible way to change a country’s growth?

Indeed, most modern historical researchers and academics do not take Diamond terribly seriously. This is primarily because of the way in which he cherrypicks examples that fit his narrative in order to create an impression of the world that wasn’t true. This ranges from him overstating the disparity between Eurasian and American technology and picking the few diseases that transmitted from animals to humans, to newer archaeological findings about the vast spread of agriculture in the American continent. His theory furthermore fails to explain the sudden and rapid growth of Western Europe during the Industrial Revolution - indeed, all Diamond offers with regards to explaining our framework is a helpful reminder that in the pre-Malthusian era, some of the growth was attributable to geographical accidents that made factor accumulation, resource allocation and technological growth a bit easier.

What is the theory of cultural determinism?

David Landes’s The Wealth and Poverty of Nations argues that rich nations got rich by developing market economies, with governments that do not interfere except to protect property rights. This arose due to the superiority of western European, and especially British culture - Protestantism gave people the freedom to think, innovate and keep the rewards of their successes. He accepts that geography had a role to play, since climate and natural resources could give countries a head start, but notes that if wealth comes too easily, people have little incentive to work hard and use their wealth productively. He points to the Spanish and Portuguese in the Americas, who gained easy wealth through gold and silver, reducing the incentive to work hard which the British had to do. The British were able to do so, because they had a culture that encouraged commerce, risk-taking and innovation, which meant it could supersede countries that were as developed, if not more so than it, during the Industrial Revolution.

However, this once again falls prey to the failure to explain countries that managed to grow after the Industrial Revolution, such as Japan or the Asian Tigers. More fundamentally, he works backwards, looking at perceived cultural characteristics in successful countries in order to extrapolate their importance, failing to explain account for countries with a similar culture that did not grow. It is therefore unsurprising that as with Diamond, the academic consensus is one that rejects culture as a definitive explanation.

What is the institutional theory of growth?

The consensus theory popularised by Daron Acemoglu and James Robinson’s Why Nations Fail is that economic institutions are the main determinant of economic growth1. This is because they determine the incentives within society - on investments in capital and technology, as well as on the allocation of resources for production. In particular, they distinguish between inclusive and extractive institutions - the former benefits everyone in the country, while the latter benefits a tiny elite at the expense of the country. Good economic institutions are inclusive ones that provide secure property rights and relatively equal access to economic resources to a cross-section of society.

Expropriation

There is a clear relationship between GDP per capita and secure property rights - however, it is difficult to make the inference that property rights cause prosperity. This is because the causal effect may be reversed or because there may be a problem of omitted variable bias - that is, there may be a third factor that causes both. In order to tease out the causal relationship, we need to find a source of variation in economic institutions that should have no effect on economic outcomes.

One would be the natural experiment of Korea. After World War Two, Korea was split into the North and the South. The North followed the model of Soviet socialism in abolishing private property, while the South maintained a system of private property. Before this separation, North and South Korea were incredibly similar - there were few geographical differences and shared the same cultural and historical backdrop. Where there were disparities, this favoured the North, with a higher concentration of industrial development. Overall, they were comparable and had approximately the same income per capita2. Following the split, they faced very divergent growth paths - while North Korea stagnated with a GDP per capita of $1,000 in 2000, South Korea had reached $16,000.

That’s only one country - does this pattern apply elsewhere?

Although this sample size of $n=1$ is unsatisfactory, a pattern of economic divergence due to differing economic institutions can be seen in the natural experiments caused by colonialism. What is found is that there are large correlations between settler mortality and colonial institutions, between colonial institutions and current institutions, and between current institutions and current per capita income3. The study suggests that European powers set up two types of colonies - for places with a high likelihood of mortality, settlers set up extractive states designed to transfer as much of the resources of the colony to the coloniser with the minimum amount of investment required. When settler mortality was low, colonisers instead settled in those places, constructing institutions with a great emphasis on property rights and checks against government power. These persisted, leading to higher per capita income in places that originally had low settler mortality.

Crucially, once this relationship has been controlled for, geographical variables provide no further explanatory power. What this suggests is that the correlations found between geographical variables such as latitude and per capita income are not direct causal links. Rather, they reflect the fact that European colonisers were only immune to the diseases that proliferated in specific latitudes - as such, geography has no causal power today, and only reflects the historical creation of institutions. However, the institutional hypothesis seems to support the cultural explanation - European settlers with different cultures set up different types of institutions. The problem with this is that as with geography, once economic institutions have been controlled for, the colonising country and religion of colonisers are not statistically significant determinants.

How do institutions differ?

The question then becomes why different places get different institutions. One explanation proposed is that societies will choose the economic institutions that are socially efficient - if the status quo is inefficient, groups will negotiate and get a more efficient set of institutions, bargaining over what share of the increased economic outcomes they get4. The problem with this is that societies would already have the most efficient institutions, rendering that an irrelevant factor, which we have seen is not the case. One response to this is to suggest that because there is uncertainty around what types of institutions are preferable, this may not result in entirely efficient outcomes - as such, only societies where they turned out to be correct will end up prosperous5. Although ideology is undoubtedly important, it seems unclear to what extent North Korea leaders would believe that their system of government is working today compared to the South. The fact that colonial empires set up different types of institutions in different places, with the British colonies in the Caribbean being significantly more extractive economies than in Canada, amplifies why ideology is an unsatisfactory explanation.

A contrasting narrative emphasises historical accidents. One theory of historical accidents argues that the differing political institutions of Britain, Germany and Russia towards democracy, fascism and communism were due to the varying types of class coalitions6. Another suggests that the origin of a country’s legal system matters - a common law system provides stronger protection of property rights compared to a civil law system7. This is reinforced by evidence that civil law jurisdictions were associated with more corruption and less judicial fairness8. Although the determination of a country’s legal system is often a historical accident and affected by which country colonised them, there is nonetheless a role for change - it is the case that countries have radically changed their political and economic institutions.

Thus the social conflict view suggests that economic institutions are determined by the actors with de jure and de facto political power within society. This is in turn affected by the existing political institutions and distribution of resources. Because political institutions are durable, and because existing elites will push for economic institutions that reinforce their power, there is a tendency for persistence and for historical accidents to perpetuate. This is made worse by commitment problems - if a dictator promises to respect property rights without relinquishing his political power, it is not a credible promise. Likewise, if a dictator agrees to transition to democracy in exchange for some compensation for the lost income, it is not a credible promise that citizens will tax themselves and transfer it to the dictator after the dictator has lost their power. The consequence is that in many cases, a stable equilibrium will be one with extractive political and economic institutions.

Thus inclusive institutions are mostly likely to arise when 1. political institutions place checks on those with political power to prevent extractive economic institutions 2. political power is in the hands of a relatively broad group with significant investment opportunities 3. there are limited rents that elites can extract from society.

But does that mean growth is ultimately predetermined by historical accidents that set up specific distributions of political power - which would lend credence to the geographical and cultural explanations? No - this is because shocks in technology and the international environment can shift this steady state. For example, property rights in Europe during the Middle Ages were initially only confined to the monarch and were designed to perpetuate their power. But as a result of the changes in the English land market and the expansion of the Atlantic trade, there was a shift in power towards landowners and merchants in 17th century England - the consequence was that they gathered enough political power to overcome the Stuart monarchs in the Civil War and Glorious Revolution, thus changing the political and economic institutions.

As such, it is clear that geography, culture and luck all play a role in the initial distribution of political power. And because power begets power, there is a certain level of persistence with respect to the institutions that exist. But this is by no means deterministic - and since various shocks can alter this equilibrium, it is institutions that decide the growth path a country faces. How can all of our theory be applied in practice? Find out in Part III!

[ economics  effortposts  ]