Using newly collected national and sub-national data, and historical case studies, this paper argues that differences in innovative capacity, captured by the density of engineers at the dawn of the Second Industrial Revolution, are important to explaining present income differences, and, in particular, the poor performance of Latin America relative to North America. This remains the case after controlling for literacy, other higher order human capital, such as lawyers, as well as demand side elements that might be confounded with engineering. The analysis then finds that agglomeration, certain geographical fundamentals, and extractive institutions such as slavery affect innovative capacity. However, a large effect associated with being a Spanish colony remains suggesting important inherited factors.