My most recent paper, which I have submitted, argues that there are problems in the way we have been using the 1851 census of Canada East. The issues relates to properly measuring volumes of grains grown and land area for farming. I argue that, since the difference in measurement units follow ethnic lines, the errors of measurements disadvantage areas that were not operating under seigneurial tenure. Thus, my assertion is that we have been using the census of 1851 to “measure away” the role that institutions might have played in economic divergence. Indeed, non-seigneurial areas exhibit higher levels of output and productivity per unit of land than when more conventional approaches to the census are used.
Here is a key passage of my paper which can be downloaded here on Academia (see table below for output measurements):
Overall, the total value of production in townships is 2.71% greater if we correct for volume measurements, and 0.22% smaller in seigneurial domains in total. The results can be seen in table 6. This may appear small, but given that McInnis and Lewis (1980) found that the productivity gap between French and English farms was somewhere between 8% and 14%, that difference is increased substantially. Another way to state the significance of these differences is that, on a per acre basis, output expands by 7.49% in townships and deflates by 1.39% in seigneurial areas. More significant is the distribution of these measurement errors. This new approach to correction implies larger errors in ethnically mixed areas – which, once again, tended to be townships.