History by the Numbers: An Introduction to Quantitative Approaches (review)

History by the Numbers: An Introduction to Quantitative Approaches. By Pat Hudson (New York, Oxford University Press, 2000) 278 pp. $74.00 cloth $19.95 paper

This book assumes "no prior knowledge of statistics or of quantitative skills" (xx). It offers an alternative to "rather more advanced" U.S. texts (xxi), but concludes each chapter with suggestions for further readings. Hudson acknowledges debts particularly to Floud, largely following his chapter divisions and coverage of techniques.¹

Hudson writes in accessible language, and not without occasional humor, even using her own youthful indiscretions to point up questionable statistical inferences. Other strengths of the work include a fairly broad coverage of useful techniques, a glossary of a dozen pages, a list of fifteen articles (mostly on British subject matter) with discussion questions, and a good deal of attention devoted to epistemological matters. Yet she appears to exaggerate the epistemological problems common to both quantitative investigation and the traditional approaches that have returned to fashion since the "linguistic turn," while giving inadequate attention to the elite bias inherent in so much of the material amenable to "thick description"—the privileging of the privileged, in effect. Similarly overlooked are recent advances (computer-assisted, if not strictly quantitative) in manuscript census indexing that help non-quantifiers to locate and obtain additional contextual information about obscure individuals for whom detailed literary sources are scarce. Hudson rightly emphasizes the need, regardless of historical approach, to engage in source criticism, but restricts this endeavor to its traditional and "social constructionist" variants, thereby neglecting to discuss opportunities and techniques for quantitative source criticism. Age heaping, for example, rates neither a glossary entry nor an index reference.

As is true with most scholars who teach and write about quantitative methods, Hudson chooses accents and examples closely related to her own research interests, namely British economic history from the seventeenth to the nineteenth century. With the exception of historical demography, social or political history and their associated techniques and measurements receive little attention. Of the dozen or so statistics that spss provides along with cross-tabulations, only chi-square and the contingency coefficient "C" are discussed. Little attention is devoted to cross-sectional regression approaches; the problem of multi-colinearity, for example, is treated only with respect to time series analysis. Although Hudson mentions Michael Anderson's work with census enumerators' books (equivalent to U.S. manuscript censuses) as an example of sampling, census public-use samples such as ipums evoke barely a glimmer on her radar screen.²

On balance, this is a welcome and useful addition to the relative short shelf of quantitative guides, but it will be of more interest to economic than social historians, to British than American or continental European specialists, and to those offering introductory or undergraduate rather than those offering advanced or graduate courses.