Banxico paper probes forced labour’s lingering effects in Ecuador

A surname’s appearance in tax records concerning colonial forced labour is correlated with its bearer’s current income, a researcher with the Bank of Mexico finds.

The paper was written by Alex Rivadeneira, who collected tax records from the 18^th and 19^th centuries and connected people’s family histories via their surnames. Rivadeneira looks at concertaje, a form of coerced labor in Colonial Ecuador.

“I find that a 10 percentage point increase in a surname’s concertaje rate reduces the current formal income of (pseudo) descendants by 1.7%.”

“Concertaje historically led to lower education levels, reduced public goods provision, increased agricultural employment, and inequality,” he says. “Concertaje also limited mobility, although its effect on immigrants is milder, suggesting migration acted as a mitigation channel.”

The author tied the farming of maize over potatoes to higher concentration of concertaje. Maize requires more labor than potatoes. He found that regions which produced maize over potatoes coerced more labour.

He finds that “an increase of 10 [percentage points] in concertaje around 1800 increased the 1990s extreme poverty rate of a locality by 4.2 [percentage points].”

Rivadeneira says the paper contributes to the debate among economists about the influence of historic institutions on economic development.

“There is little evidence about the persistence of institutions’ effects on individuals, particularly regarding intergenerational impacts on the descendants of those who lived under these institutions,” he writes. Concertaje survived in some form until 1964, he notes.

The author says that although his paper focuses on forced labour in Ecuador, the mechanics are similar to those used throughout Latin America. “Furthermore, forced labor has been the most common type of labor relation throughout history,” he writes. “However, empirical evidence about its effects is scarce due to data limitations.”

The author checked the robustness of his findings against control samples. “Placebo regressions in which surnames are randomly reshuffled show neither statistically nor economically significant effects,” he says. “Instead, weighting for surname rarity shows larger ones.”