Principal investigator: Angelo Cruz do Nascimento Varella
Article title: UNDERSTANDING MECHANISMS UNDERLYING PEER EFFECTS: EVIDENCE FROM A FIELD EXPERIMENT ON FINANCIAL DECISIONS
Article authors: Leonardo Bursztyn, Florian Ederer Bruno Ferman and Noam Yuchtman
Location of the intervention: Brazil
Sample size: 150 investor pairs
Main theme: Finances
Type of Intervention: Experiment on investment decisions
Main Variable of Interest: Social influence
Evaluation method: Experimental Evaluation (RCT)
Policy Problem
Frequently, the choices of relatives, friends, and professionals we admire influence our own decisions. This phenomenon, called social influence, affects our preferences for products and services, having a direct impact on our lives and society as a whole. However, the choices of peers are not always the best options, and in these cases, the influence that such choices have on individual decisions generates undesirable consequences, both individually and collectively.
These social consequences can be observed in various sectors, such as public health and commerce, but they have a particularly worrying impact on the financial market. This is because social influence is linked to herd behavior and other market imperfections that generate excessive volatility and add uncertainty.
Assessment Context
Despite the understanding that peer influence affects individual decisions, it is still necessary to investigate how this phenomenon occurs, specifically which channels of social influence impact investors' financial decisions. This research identifies two channels through which social influence can occur.
Firstly, social learning is characterized by the effect that a peer's interest in acquiring an asset has on a person's decisions. Basically, it is the individual's conclusion that the other person's desire to obtain a good or asset is derived from a positive characteristic of that item.
Secondly, social utility is defined by the influence resulting from the possession of an asset by a peer; that is, the indication of possession of a good also has effects on an individual's decision to acquire or not acquire the item in question.
Obviously, both channels are amplified if individuals perceive characteristics of experience or expertise in their peers.
Policy Details
In order to investigate and measure the social influence on Brazilian investors, researchers worked together with a large financial brokerage firm in 2012 to create an experiment that could assess the social influence of both learning and utility channels on investors' final decision-making.
Thus, the researchers offered a new financial asset to 150 pairs of investors who had a close personal relationship (friends and family) and who were clients of the partner brokerage firm. The aim was to offer a high-risk asset with a minimum investment of 2 reais, which represented about half of the average monthly income of the selected investors.
Each pair of investors was randomly assigned as Investor 1 and Investor 2. The first investor was given the chance to invest in the asset without any information about the second individual. The second investor, however, was provided with different types of information, depending on the draw.
Assessment Method
There is an inherent difficulty in differentiating channels of social influence, since the interest in acquiring and the acquisition itself become confused in empirical analyses. Thus, the researchers used lotteries to make the available information random and to separate each type of social influence in an effective way.
Upon contacting Investor 1, the researchers stated that the investment was scarce and randomly selected the possibility of the individual actually acquiring the asset. Then, they conducted another random selection and contacted Investor 2. Based on the result of this new selection, the second investor was informed of the first investor's intention or successful acquisition of the asset.
If the second investor knew that the first investor intended to buy the asset but was unable to, there was a learning influence channel without social utility. If, on the other hand, they knew of the successful acquisition by the first investor, there was a situation of learning and social utility, thus separating the two conditions into isolated situations.
Results
The results obtained point to positive effects derived from both channels of social influence. It was also observed that the effects were greater in investors with less financial sophistication, a measure extracted from a questionnaire applied after the experiment. In other words, investors with less experience and financial expertise tend to be more influenced by social factors.
With regard to social learning, when Investors 2 learned that Investors 1 wanted to acquire the asset, the intention to buy was positive on 71% of occasions, a considerably higher result than the group of Investors 2 who did not receive such information and were interested in the asset in only 42% of opportunities.
In situations where Investors 2 received information that Investors 1 had in fact acquired the asset, creating a learning situation and a social utility, Investors 2 opted to acquire the asset in 93% of cases.
Lessons in Public Policy
The results demonstrate the importance and magnitude of social influence on the decisions of Brazilian investors. This is relevant because it shows that social factors are decisive in the individual decision-making process and, consequently, affect society as a whole. Not only does social influence directly impact the financial market, but the research findings are also important for the formulation of public policies capable of exploiting these characteristics.
A prime example is public health, such as vaccination campaigns and the promotion of healthy habits, like physical exercise and a balanced diet. By exploring channels of social influence, it's possible to maximize campaign results, for example, and encourage good collective practices. Experiments in finance are useful in this regard, as they allow for the exploration of psychological axioms and enable a better understanding of human nature.
Reference
BURSZTYN, Leonardo et al. Understanding mechanisms underlying peer effects: Evidence from a field experiment on financial decisions. Econometrica, vol. 82, no. 4, p. 1273-1301, 2014.
