Increasing returns to scale in the towns of early Tudor England

Rudolf Cesaretti, José Lobo, Luis M.A. Bettencourt, Michael E. Smith

Research output: Contribution to journalArticlepeer-review


Urban agglomeration economies make cities central to theories of modern economic growth. There is historical evidence for the presence of Smithian growth and agglomeration effects in English towns c.1450-1670, but seminal assessments deny the presence of agglomeration effects and productivity gains to Early Modern English towns. This study evaluates the presence of increasing returns to scale (IRS) in aggregate urban economic outputs—the empirical signature of feedbacks between Smithian growth and agglomeration effects—among the towns of 16th century England. To do so, we test a model from settlement scaling theory against the 1524/5 Lay Subsidy returns. Analysis of these data indicates that Tudor towns exhibited IRS—a finding that is robust to alternative interpretations of the data. IRS holds even for the smallest towns in our sample, suggesting the absence of town size thresholds for the emergence of agglomeration effects. Spatial patterning of scaling residuals further suggests regional demand-side interactions with Smithian-agglomeration feedbacks. These findings suggest the presence of agglomeration effects and Smithian growth in pre-industrial English towns. This begs us to reconsider the economic performance of Early Modern English towns, and suggests that the qualitative economic dynamics of contemporary cities may be applicable to premodern settlements in general.

Original languageEnglish (US)
Pages (from-to)147-165
Number of pages19
JournalHistorical Methods
Issue number3
StatePublished - Jul 2 2020


  • 1524/5 Lay Subsidy
  • Early Modern Europe
  • Settlement scaling
  • Smithian growth
  • agglomeration economies
  • urban economics

ASJC Scopus subject areas

  • History


Dive into the research topics of 'Increasing returns to scale in the towns of early Tudor England'. Together they form a unique fingerprint.

Cite this