--- - |2 Abstract This article argues that the 1851 census of Canada East (the modern‐day province of Quebec) requires a set of important corrections. Using corrections based on ethnic origin composition, I demonstrate how significantly wheat and oat yields were underestimated in Canada East. More importantly, I argue that the measurement errors are not randomly distributed and that they are biased against attempts to test the role of institutions. I show how the new method of correcting the data change our interpretation of agricultural efficiency in Lower Canada in the mid‐19th century. While this correction may seem minor, it shows that in the past, the data took a form that was biased against numerous hypotheses concerning land tenure institutions. - 'Social Science Quarterly, EarlyView. '