Mainstream economists should elevate the status of "empirical rigor" to match that of "mathematical rigor" and focus on the alignment of theoretical models with socioeconomic realities. Photo: TUCHONG

Since the evolution of economics into a “modeling science” in the latter half of the 20th century, critics have repeatedly argued that “economic analysis has become detached from reality due to excessive reliance on technical tools, rendering it ineffective in addressing major social and policy issues.” The 2007-2009 Great Recession further intensified such criticisms. Mainstream economists, under mounting scrutiny, began to recognize that the overemphasis on logical “rigor” had come at the expense of real-world “relevance.”

Prior to the 1940s, the trade-off between “rigor” and “relevance” did not pose a substantial challenge to economists. Particularly during the two World Wars, as many economists prioritized empirical research and historical analysis, the emphasis on “empirical rigor” lent economic theory a strong realist character, intrinsically unifying “rigor” and “relevance” through a focus on empirical data.

The “formalism revolution” of the 1950s, however, established the position of mathematical language in economics, elevating “mathematical rigor” to a primary goal of economic research. A crack in the unity of “rigor” and “relevance” thus emerged, although it remained largely unnoticed then. By the 1970s, growing social issues raised concerns among economists about the discipline’s excessive allocation of intellectual resources to the pursuit of “mathematical rigor” at the cost of “relevance,” sparking a wave of criticism that lasted for three decades.

The crux of the debate over “rigor” vs. “relevance” lies in the transformation of the concept of “rigor,” reflecting the evolving perspectives on (natural) scientific knowledge. Before the 1950s, “rigor” in economics was primarily understood as “empirical rigor,” aligning with the understanding of “rigor” in the natural sciences—particularly in physics and mathematics—where rigorous mathematical modeling was expected to be directly and explicitly grounded in empirical evidence.

However, following the crises in physics and mathematics in the early 20th century, mathematics came to be viewed as a practice focused on constructing systems consisting based on axiomatic sets and their deductions. Consequently, from the 1950s onward, “rigor” in the natural sciences became increasingly associated with “mathematical rigor,” fueling an enthusiastic pursuit of formalist models and axiomatic theories. Influenced by this trend, since the mid-20th century, economic modeling gradually shifted its focus toward the logical consistency of theories rather than establishing robust links between theory and data.

Admittedly, economics as a discipline has developed its own set of relatively independent academic standards that represent the mainstream disciplinary position. Nevertheless, external factors were equally—if not more—important in the prioritization of “mathematical rigor” over “empirical rigor.” Governments and large foundations, as important sponsors of economic research, played a key role, while societal and public demands provided strong impetus.

More importantly, the standard of “mathematical rigor” became increasingly institutionalized, contributing to the development of an academic hierarchy and evaluation system favoring mathematical formalism in publication, consulting, education, and employment. This institutionalization further entrenched a narrow definition of “mathematical rigor,” whereby only studies adopting mainstream mathematical formulations and modeling methods were recognized as legitimate economic research, exacerbating the “trade-off assumption.”

In the 21st century, criticism of economics has intensified, underscoring the necessity of overcoming the constraints imposed by narrowly defined “mathematical rigor.” A key lesson from the history of economic thought is that the current dominance of “mathematical rigor” is not the result of natural or inevitable processes. To address the dilemma caused by overemphasizing “mathematical rigor,” the status of “empirical rigor” should be elevated to match that of “mathematical rigor” within the theoretical evaluation framework.

Mainstream economists should embrace the development of non-mainstream mathematical modeling approaches and create greater space for non-mathematical forms of “rigor.” They should also rigorously define core concepts and focus on the alignment of theoretical models with socioeconomic realities.

Li Lili (professor) and Ren Weiwei are from the School of Economics at Renmin University of China.

Edited by WANG YOURAN