The history of economics is compelling for it positions the discipline within large social and cultural narratives. As a cognate discipline of economics, the history of econometrics has become an established field of research in its own right and historical narratives are expanding rapidly. The heterogeneity in economic referents that econometrics must factor into model construction had led, increasingly, to tensions regarding the econometric reconciliation of the empirical and theoretical in an age of rapid technological advancement. Resultantly, the developments in academic exercise and practical application that A History of Econometrics illuminates is especially apropos to those in the social sciences.
Qin’s account details the history of reformations in econometrics since 1970, largely focusing on the reigning Cowle’s Commission (CC) approach in structural econometrics and the movements that have either directly challenged or sought to extend its methodological foundations. Moving chronologically from the historical emergence of the CC paradigm, Qin details the rise of competing methodological schools and provides explanatory case studies before returning to a detailed citation analysis that explains the continued reign of the CC paradigm in the face of such reformatory efforts. Expertly weaved throughout the chapters are narratives that illuminate the interdisciplinary tensions between theoretical and empirical methods and the difficulties of corresponding econometric models with their real world referents.
Chapter One is an explanatory chapter detailing the historical emergence and consolidation of the CC paradigm. Chapter Two is dedicated to the rise of Learner’s Bayesian regression model selection strategy and its attempts to remedy the empirical fragility of the CC’s a priori model specifications. Chapter Three discusses the emergence of the vector autoregressive approach (VAR) as an internal reformatory movement within the larger CC tradition. Chapter Four describes the London School of Economics’ (LSE) developments as ones that shifted the criterion of model selection away from the CC’s a priori fixed parameters to data-instigated parameters. Chapters Five and Six offer meticulously detailed case studies on the Phillips Curve and business cycles to illuminate the issues of navigating theoretical–empirical trade-offs in relation to model choice. Chapters Seven, Eight, and Nine further define these tensions by exploring the variegated approaches within econometrics to define appropriate error terms, parameters, and selection strategies. Chapter Nine is the most theoretically fruitful component in the text as it addresses the problematic gap between theory-driven and data-driven methodologies and the resulting discrepancies between econometric models and their real-world referents. Chapter Ten concludes with a citation analysis that justifies the success of the CC paradigm in terms of research emulation and subject diffusion. Despite the divergence within the CC paradigm, there has been no consensus on a clear methodological replacement for it. Thus, as the author concludes, no Kuhnian paradigm shift has yet occurred.
Two narratives within A History of Econometrics text may be discerned: the first concerns the interdisciplinary tensions in constructing plausible links between economic rhetoric and statistical data-sets, and the second illuminates the problematic break between econometric models and their real world referents. Qin constructs the first narrative by examining three major methodological approaches (Bayesian, LSE, and VAR) to resolve issues in model choice and describing the somewhat crescive history of attempts to define parameters, error terms, and selection. For example, the Bayesian approach never fully developed into a successful alternative approach over the traditional CC methodology for it was believed to be overly subjective in its statistical dealings with issues in model selection. In attempting to methodologically address the gap between macroeconomic theory and econometric data, VAR defined the desired statistical properties of residual error terms. This development in empirical measurement lessened the reliance on economists to provide accurate a priori structural parameters. Moreover, in making data-based criterion and predictive ability fundamental elements of model selection, the LSE approach propelled the discipline toward data-instigated instead of a priori set methodologies. Evidently, economic theory illuminates obfuscated relationships within data-sets, and data-sets are necessary to justify theoretical parameters: each implicates the other, that much is clear, but to what extent? Qin thoroughly constructs how attempts to formalize this relationship between theoretical and empirical methods remain considerably unresolved despite the historic shift toward data-instigated methods.
Despite technological advances that have raised the mathematical objectivity of the discipline, A History of Econometrics notes that academic econometrics still has a problematic relationship with reality for its mathematical laws remain disassociated from their ‘real’ material referents. The dissociation between practice and reality is most clearly illuminated in Qin’s discussions concerning the limited predictive capabilities of econometric models. Until new methods appear that lessen the dissociation between the academic practice of econometrics and its material referents, the structural break between the discipline and its real-world applicability remains and the CC paradigm will likely continue its reign.
Qin’s work offers an excellent historical construction of post-1970 developments in econometrics by emphasizing the dominance of the CC paradigm, the discipline’s unresolved mediation between economic theory and data-sets, and the oft-problematic rupture between academic models and their real-world applications. Attempts at formalizing an appropriate trade-off between theory and data are trending toward data-instigated methods, but evidence of the practical ‘real-world’ benefits of this approach is wanting. Not only does A History of Econometrics clearly untangle the history of these intra-disciplinary tensions, it demonstrates Qin’s expertise in relating these tensions to larger philosophical questions such as the correspondence between the resulting econometric models and their material referents.
If there is one critique, it must be noted that the role of technology and its relation to econometric innovation remains largely unaddressed throughout the text. Qin routinely points to issues surrounding the of lack of computational abilities that have hindered methodological developments, such as the ‘technical curse’ that hindered the VAR’s forecasting abilities and the lack of computation abilities that would have assisted the early Bayesian statistical regression analyses in becoming a more meaningful reformatory movement, yet never goes further to demonstrate how technological developments may have guided the diffusion of research topics within the discipline itself. It is clear that the reformatory movements since 1970 concomitants the data discovery with cognitive processes such as interpretation. These cognitive processes have been embedded into the technologies of the discipline, for cognitively derived classification systems now command the data analysis processes. It would be fair to say that the discipline of econometrics can no longer be removed from its technological equipment and this narrative is something that warrants critical attention. While A History of Econometrics explains that the shift to data-driven regimes was mirrored by the computerization of model choice methodology, the lack of formal discussion regarding intersections between technological advancements and econometric innovation leaves the reader wanting.
There are also questions concerning the molding of recalcitrant data into econometric models that plausibly correspond with their material, real-world referents. While Qin intricately details the discipline’s issues in formalizing empirical–theoretical trade-offs, she reserves only the briefest commentary for addressing the problematic relationship between abstraction in models and their correspondence with their underlying material reality. Qin refers to the hermeneutics of data and philosophies of science, yet does not clearly explain how such approaches may correspond to her historical account of econometrics. In this regard, A History of Econometrics shies away from offering any substantial theoretical discussions concerning the ontological nature of econometric materializations, though this may well be argued to travel beyond the text’s intended purpose.
A History of Econometrics is meticulously researched, thoughtfully arranged, and is an excellent text for those seeking to integrate the history of econometrics into larger narratives within the social sciences. It would best serve the reader looking to supplement an already strong understanding of the discipline, for the text reads like a highly technical blueprint of econometric innovation and assumes a specialist audience. For the price, it provides a near-overwhelming amount of historical detail and will satisfy any scholar with an interest in the field.