Has Behavioural Economics made Economics less insular?

Graphic painting telepathy. Understanding. Empathy.By Alexandre Truc

Postdoctoral Researcher, Université du Québec à Montréal (UQAM), Centre interuniversitaire de recherche sur la science et la technologie (CIRST)

 

Economics is a discipline often described as “separate” (Hausman, 1992) or “insular” (Fourcade et al., 2015) from other social sciences. When Daniel Kahneman – a psychologist – won the Nobel in 2002, many hoped that behavioural economics (BE) would lead to more interdisciplinary research in economics. However, multiple small-scale studies (Grüne-Yanoff, 2016:p357) pointed out that interdisciplinarity in BE has been decreasing since its development as a discipline (Braesemann, 2019), or that it is at best limited (Sent, 2004). Is behavioural economics becoming less interdisciplinary? To answer this question, we mapped and measured the interdisciplinary practices of behavioral economics using 5423 articles.

A first challenge to study the interdisciplinary practices of behavioural economists is to identify articles in the field. Behavioural economics cannot be reduced to a set of journals, authors or institutions, especially after the field became very successful in the 1990s and 2000s. For the present article, we chose to use a four step approach (Figure 1) combining data from Clarivate’s Web of Science (WoS) and EconLit.

Figure 1
Fig 1: Methodology used to identify behavioural economics

STEP 1 – JEL corpus: we identify all articles considered as behavioural economics in EconLit based on the following JEL codes: Micro-Based Behavioural Economics, Macro-Based Behavioural Economics, Behavioural Finance, Neuroeconomics.

STEP 2 – CLASSICS corpus: we identified the classics of our JEL corpus by taking all articles cited by at least 8 articles of our JEL corpus.

STEP 3 – EXTENDED corpus: we took all articles citing at least 8 articles of our CLASSICS, or cited by at least 6 articles of our CLASSICS.

STEP 4 – Final corpus: CLASSICS + EXTENDED (5 423 articles)

Using this corpus, we produced a direct-citation network, and applied Leiden modularity based algorithm to identify the main communities of BE (Figure 2):

Figure 2
Fig 2: A Map of behavioral economics (color coded by communities; nodes sizes were scaled to the number of citations in the network)

The publications of behavioural economists transformed rapidly since the field emerged in the 1980s. While in the early years, around 40% of BE’s articles were in psychology, this proportion decreased to less than 15% by the 2010s (Figure 3). The decreasing importance of psychology was not the result of the rise of economics. Instead, we find a rise of management (finance, consumer research, and marketing), as well as a variety of other disciplines. This includes a slow and general increase of general sciences (ScienceNature…), as well as spikes of interest for a few disciplines like law in the 2000s and neurosciences in the 2010s.

Figure 3
Fig 3: Discipline of the articles of the corpus

We find major differences in the disciplinary composition of BE’s specialties (Table 1): Behavioural Finance is mostly management and economics; Neuroeconomics mostly neurosciences, psychology and general sciences. The majority of the remainder are mostly economics and management with psychology as a tertiary discipline.

Table1
Table 1: Summary of the clusters

We find two exceptions:

First, among the three bounds of rationality (time, risk, social), Social Preferences is the only cluster with almost no psychology.

Second, the Biases, Heuristics and Nudges cluster is mostly psychology, with economics as a secondary discipline. This is the cluster where we find most of the (older) foundational psychological articles the network rely on.

The relative importance of these specialties over time influenced the disciplinary composition of the field (Figure 4). In the 1980s, BE was mainly about Biases, Heuristics, Risk and Uncertainty. Two clusters with Kahneman and Tversky at their centre, and with a strong interest in psychology. By the 2010s, Social Preferences – one of the less interdisciplinary clusters – became the dominant topic of the field.

Figure 4
Fig 4: Share of communities over time (smoothed using three years moving average and local polynomial regression)

Beside disciplinary composition, another way to think about interdisciplinarity is to look at the interdisciplinarity of economics articles within BE. Do they cite more psychology? The short answer is “no”. We find less psychology, more management, more of a variety of other disciplines; but not more economics (Figure 5). The decreasing importance of psychology is not a return of economics to a purely self-contained discipline. Instead, economics articles in BE show an interest for a greater diversity of disciplines, even if the importance of psychology decreased.

Figure 5
Fig 5: Share of disciplines in economics references (smoothed using three years moving average and local polynomial regression)

Overall, since the 1980s, we find less interactions with psychology in BE. However, there is a twist. The interdisciplinarity between economics and psychology did not actually decrease. If we look at the references of all economics articles, BE or not, we do find an increasing share of references to psychology (Figure 6). In the 2010s, the share of citations from economics to psychology rose to a level similar to the 1960s with around 1.6% of all economics’ references to psychology. Moreover, despite the decreasing importance of psychology in BE, in 2018, economics articles in BE were much more open to psychology than the average article in economics  (10% versus 1.6%).

Figure 6
Fig 6: Share of disciplines in the references of all articles published in economics

 

Behavioural economics acted as a bridge building program: as BE became more popular, the use of psychological references became more widespread in economics even if contemporary individual BE’s articles cite less psychology than in the 1980s. In the 1980s, behavioural economics was just a handful of individuals with strong ties to psychology (e.g., Kahneman, Tversky). They published in both disciplines, and economics articles had many references to psychology. As the field became successful in the 1990s, and even more so in the 2000s, the number of behavioural economics’ articles increased rapidly (Geiger, 2015); and individual BE articles became less intensely interdisciplinary. The decreasing intensity in interdisciplinarity was compensated for by the increasing number and importance of BE articles in economics. The overall level of interaction between economics and psychology (as well as general sciences) increased very rapidly in economics as BE became more influential.

 

The full research paper by Alexandre Truc “Mapping behavioral economics and its interdisciplinary practices” (May 2020) is available at: https://drive.google.com/file/d/1BmAtpbW0efAIiVLtycgTgieAF01H6UWg/view