Human Capital Flashcards
Rocha et al 2017
Effect of state-sponsored settlements in Sao Paolo, channeling high-skill immigrants to specific settlements.
LR effect of 15% higher incomes in these areas by 2000 (and Sao Paolo persists as the dominant region). The municipalities had higher literacy rates (33% higher in 1920) persistently more school enrolment, and greater involvement in industrialisation. Also attracted other educated individuals.
Settlers brought social capital like trust, and municipalities with a settlement had more expansive schooling systems over time.
Not too persuasive - doesn’t show education as fundamental, only important (obviously). Does demonstrate persistence and shock effects.
Glaeser et al 2004
Institutions or Human Capital?
Argues institutions do not drive growth, but the human capital behind them does. Autocrats can succeed through good policies, but education is needed to sustain democracy; hence it tends to come after autocracies achieve growth (Taiwan, South Korea). Relate to Acemoglu Coase Theorem.
Acemoglu’s reversal: settlers brought a lot of human capital, knowledge, etc. Related to institutions and hard to separate.
Evidence - 1960 education is a strong predictor of subsequent growth by 2000. Human capital leads to growth through innovation, and more benign/less violent politics, and “good” institutions.
Uses the example of NK and SK both being governed by dictators, but with one still utilising capitalism and developing human capital. This drove the growth of better institutions. It is therefore hard to measure institutions independent of outcomes (they are endogenous, with HC being the common root factor). Response would be that while political institutions were nondemocratic, the economic constraints were more growth friendly (eg little appropriation).
Squicciarini and Voigtlander 2014
Human Capital and Industrialisation
Considers the importance not of mere literacy and education (needed to use technology, SR growth), but upper tail knowledge (“knowledge elites”) (needed for innovation, LR growth). Proxies by subscriber density to the Encyclopedie in France during the Enlightenment.
Subscriber density (controlled for effects of wealth density) correlates with city growth (output, employment, city population) in the long run - 1750-1850, but not 1700-1750. Indicates it became significant during the IR.
Relates to the Huguenots, and how their presence in a city correlated to growth, through their emphasis on education and science. H density in 1670 correlates with subscriber density in 1750. This impact has not been as persistent as Sao Paolo; in Brazil inequalities have persisted along HC lines, while in Europe they exist but not along the same lines.
De La Croix et al 2017
Guilds versus Clans
Compares models of sharing technology and training in China (eg) and in Europe. Families and clans restrict access to knowledge to outsiders and create inefficiency, less innovation. Guild and apprenticeship system allowed spreading across Europe (particularly journeymen). More dissemination enables innovations to boost productivity more.
Training system enabled by guilds controlling the industry to enforce contracts (but would also be inefficient, constraining trainees to keep wages high). Guilds are always more efficient than clans. In China, clans held societal significance; this equilibrium was hard to move away from (while Europe was nuclear).
Adopting guilds becomes more feasible as population density rises. Europe started from a family equilibrium, hence the shift was easier as the equilibrium was in itself less efficient. Whereas the clan equilibrium was better at first, but became entrenched and less efficient.
Comin, Easterly, and Gong 2010
Long-term human capital persistence
Technology variation in 1500 explains 18% of modern logGDPpc variation. Other evidence for technology in 1000BC and 0AD.
But this seems unpersuasive; China and the Islamic world were well ahead then. Hard to disentangle the effects from European status.
Explanatory model posits that the cost of adopting new tech falls with the current level of technology; cumulative growth. Eg, since tech can be repurposed into new ones, economies of scale, spillovers from sectors, etc.
Easterly and Levine 2012
AJR emphasise institutions as the key channel, while GLLS emphasise human capital. EL find that most important determinant of outcomes is share of europeans during colonisation (not now), implying the effect was the diffusion of human capital. Also undermines the RoF thesis, since at all levels of settlement, more settlement is associated with more modern income (rather than none to few implying extraction and bad outcomes).
Finds indigenous, not settler, mortality key for settlement (contra AJR). Effect of euro settlers disappears when we control for measures of HC or institutions; hard to isolate, but GLLS argue HC is what then determines institutions, since education is required to support them. British coloniser identity insignificant.
Key finding is that the RoF seems less agreeable, but AJR and GLLS are both right that institutions and HC are channels (more evidence to GLLS, since marginal effect of settlement stronger at these lower levels).
