Essays on the the Mathematization of Economics

The thesis takes as its starting point the alleged “founder” of “scientific” economics, Lionel Robbins, and investigates how economics has evolved in time, detaching itself ever more from the richer texture of Robbins’ view which was more akin to classical political economy. I argue that the top journals of economics were key players in the significant epistemological changes the discipline underwent after the 1940s, first with the increasing mathematical formalization, and later with the rise of empiricism in an attempt to mimic the natural sciences. If the 1940s and 1950s marked the rise of formalization in economics, there has been a more recent shift from the Cowles econometric approach to quasi-experimental methods, which have grown considerably in importance since 1990. This shift originated with the disarray in which the simultaneous equations approach fell during the 1980s, with its contested reliance on economic theories based on the optimization of agents or firms. If economics has certainly been relying less on theoretical models and becoming more data-driven in recent decades, hence more applicable to policy matters, this does not necessarily mean that political economy is resurfacing. Economics may be turning into a more useful instrument for designing public policies, but it remains to be seen whether the social, political and institutional elements that were extruded in the transition from political economy to economics will have any important role to play in the new era ofdata-driven economics. Chapter one argues that parallel to the mathematization of economics there was a redefinition of the methodology of economics. In order to do so I assess the influence of British classical economists on Lionel Robbins in what concerns the demarcation between the epistemology of the natural and the social sciences. With the mathematization of economics this distinction was displaced and economists increasingly started to mimic the methods of natural sciences, leading to the dehistorization and desocialization of economics. Lionel Robbins provides an interesting window into investigations of the changing nature of economics not only because he was concerned about the methodology of economics, but also because he wrote in a period when economics was quickly becoming more quantitative. In this sense the lingering influence of British classical economists on Lionel Robbins shows that the mathematization of economics after the 1930s happened in tandem with an epistemological change whereby economists increasingly endorsed the methods of the natural sciences as the proper way to approach their subject matter. In recent years the role of top journals of economics in molding research has been widely debated and the chapters 2 and 3 of the thesis contribute to this debate by examining how the language used by economists evolved since the foundation of Econometrica in 1933 until recent years. The second chapter focuses on the consolidation of mathematical economics between the 1930s and the 1960s. I argue that one of the drivers of the mathematization of economics was the ideological context of the period. As a response to totalitarian tendencies taking place in Europe, quantitative methods grew in importance because they were perceived as being value-free. Using latent semantic analysis I show that there was a convergence in the language used by the top journals of economics and in this sense the origins of the so-called top five journals can be traced back to the 1960s when they endorsed the methods championed by the Econometric Society. Chapter three then investigates the role of the American Economic Review, the Journal of Political Economy, and the Quarterly Journal of Economics in the development of mathematical economics. Using bibliometrics it is shown that the papers published in these journals are strikingly similar between 1940 and 2010. In this sense I argue that mathematical economics was increasingly incorporated by the top general journals and the rest of the profession followed suit, not least because the creation of citation metrics in the 1960s created a hierarchy within