In 1086 William the Conqueror carried out an extraordinarily detailed survey of the wealth of England. His purposes were controversial then and have remained so; it would be far easier to interpret the data which Domesday Book contains if we could be certain we clearly understood the purposes for which it was collected. For most estates across most of England, however, Domesday records three critical pieces of specifically economic information: the tax assessment assigned to each manor (expressed in hides); the value (valet) of the manor, which was sometimes lower than the rent (reddit) being collected from it in 1086; and the principal resources which made up its value. Using modem econometric and statistical techniques, MeDonald and Snooks have analyzed the relationship between these economic variables in two Domesday counties, Essex and Wiltshire. Having established the reliability of the Domesday statistics, they have then devised an econometric model of manorial production in late eleventh century England.

This is an interesting and provocative undertaking, which promises to produce several further volumes of results beyond the present one. Here the authors are principally concerned to explain their methods and vindicate their models. Previous studies of Domesday statistics, the authors argue, have taken a descriptive rather than an analytic approach, mapping the distribution of resources or the average valuation of manors without attempting to analyze the economic relationship between valuations, assessments, and resources. When such analysis has been attempted, moreover, the results have been vitiated, in the authors' judgment, by non random sampling techniques, a concentration on deviant over typical examples, and a set of counter intuitive economic assumptions, most notably the claim that tax assessments were arbitrarily imposed and therefore bore no relationship to the economic value of the manors. Using exceptionally sophisticated statistical methods, the authors are able to avoid the errors they ascribe to their predecessors, and to show impressively high correlations between assessments and valuations (0.70), and between valuations and recorded resources (0.90). Using their production model, they are also able to suggest some tentative conclusions about the substitutability of one manorial resource for another, and the degree to which Domesday manors benefited from economies of scale.

. . .

McDonald and Snooks have . . . made a valuable contribution to both groups with this book. By dividing the book in half, and relegating their statistical methods to part II they have made it possible for any historian, regardless of training, to follow their arguments and conclusions. In part II, they have developed the mathematics behind their models with admirable clarity, providing something very like a textbook in historical econometrics. These are valuable methods to have added to Domesday studies, and while they will not resolve all the problems this document poses, they may well allow us to resolve some longstanding Domesday conundrums which have defied resolution by more "conventional” means. This is an exciting prospect; I for one will look forward to future volumes.

-- Robert C. Stacey, Journal of Economic History, XLVII (3), September 1987, pp. 780–1.

Return to top

This book follows up several well known articles by the authors which were a pioneering foray beyond the chronological boundaries within which quantitative economic historians normally operate. In fact, the main addition to their earlier publications is in the form of three chapters on methods intended both to spell out some basics of statistics for the cliometric novice and to exposit some quite advanced material on production functions for the aficionado. The authors' style is hard hitting and there is a significant amount of missionary zeal in their advocacy of the virtues of the cliometric approach relative to that of traditional historians of Domesday Book.

McDonald and Snooks' work has many merits. The econometric analysis is admirably executed and clearly laid out and they show that the relationship of tax assessments to recorded values of manors and also of recorded values to resources employed was in each case quite well defined. These results are important and will surely lead to widespread revision of contrary views long held by scholars in the field.

. . . McDonald and Snooks do see their work as establishing an agenda for further research rather than as a definitive study.

. . . I enjoyed reading this book and I learnt a lot from it. McDonald and Snooks have made an important contribution and have opened up some interesting avenues for further research while demonstrating the value of modern quantitative methods in making sense of apparently chaotic data.