In the 1970s, Goldschmidt found that industrialization of swine production in Iowa resulted in declining social and economic returns to neighbouring farming communities. This finding is confirmed by several more recent studies in the United States, indicating that regions dominated by family farms possess better socioeconomic conditions compared to regions with larger farms. This paper explores the Goldschmidt hypothesis in the Canadian context with data from the Census of Canada and the Census of Agriculture (2006) at the level of the Census Consolidated Subdivision. Measures of industrialization include indices of farm capitalization and farm receipts, and a ratio of total pigs and total cows per farm within a region and indicators for socioeconomic status include average income and a poverty rate. Bivariate and multivariate statistics show that the relationship between agricultural structure and socioeconomic outcomes is often weak, or potentially non-linear. The mean number of pigs per farm in a Subdivision, for example, is associated with higher average incomes to a point (approximately 2500 pigs) and then, on average, no further income gains are realized from larger herds. Based on these findings, we reject the Goldschmidt hypothesis and construct a more complex picture of the social effects of agricultural industrialization in rural Canada.