briq Newsroom

beliefs

How associative memory shapes belief formation

Belief-based decisions are not only relevant for individual economic outcomes but may also have important effects on the economy as a whole. This is why belief formation is an important area of study for economists. briq has supported pioneering research on this topic, for example through workshops with leading international scholars (see the 2019 report and the program of the 2021 “Beliefs” workshop which took place in early May) and an ERC-supported project led by briq Research Director Florian Zimmermann.

In a video produced by Latest Thinking, Zimmermann explores the role that memory plays in belief formation. In particular, he studies associative recall, the way in which the recall of memory trace is triggered by cues that are mentally associated with that trace.

His experiment involves participants being provided with information (positive and negative) about hypothetical companies. After a time lapse, they are asked to assess the values of these companies. The research shows that respondents consistently overreact to the last piece of information that they receive. It also presents a new experimental paradigm for studying the role that associative recall and other memory imperfections play in the process of belief formation.

Associative memory and belief formation

In a recent episode of “Behavioural Science Uncovered”, a podcast with a particular focus on the research process, briq Research Director Florian Zimmermann talks about his paper “Associative Memory and Belief Formation,” co-authored with Benjamin Enke and Frederik Schwerter.

Florian Zimmermann

The paper experimentally investigates the idea that people are more likely to recollect items that are cued by current context. This is because (i) people do not constantly have access to their beliefs so they may need to reconstruct prior information from memory; (ii) similar real-world news are often embedded in similar memorable contexts. The paper finds a predictable and quantitatively meaningful role of associative memory in the formation of beliefs.

To learn more about the paper and the origins of the project, but also about Florian’s experience preparing an ERC starting grant application, and what his typical working day looks like, tune in to the podcast at bsuncovered.com!



briq student fellow portraits: Peter Andre

The briq scholarship program supports about 25 Ph.D. students at the Bonn Graduate School of Economics. Some of them are directly involved in briq research activities. In a series of short portraits, the briq newsroom introduces our student fellows and what they do. One of them is Peter Andre, who has been at briq since October 2016.

What kind of topics are you working on?

My current research focuses mostly on beliefs and perception. Beliefs matter not only for decision-making under risk as traditional economics would claim. They can matter whenever behavior and reasoning depend on the environment – which is to say: virtually always. It’s usually our perception of the environment, our mental representations, that shape our thoughts and conduct. Now, in some cases, it may be reasonable to assume that our perceptions are correct and reflect what’s out there. However, we know already that this assumption will lead us astray in many important settings: educational or financial decisions, parenting, causal or political reasoning, and social or moral behavior.

Can you give an example?

In a project with Chris Roth, Johannes Wohlfart, and Carlo Pizzinelli, we study macroeconomic belief formation. We analyze how unemployment and inflation expectations react to hypothetical exogenous shocks such as changes in the oil price or federal funds rate. These beliefs are important because they can determine the trajectory of the economy and the effectiveness of policies. They can fuel booms and busts. So are they formed in line with standard macroeconomic models? It turns out that forecasts are often off the mark. This holds particularly for inflation forecasts and respondents with little knowledge and exposure to education.

What is a typical problem you are facing in your research?

In the context of beliefs, measurement is a huge challenge. You often face a trade-off between making the elicitation technique easy to understand for the respondent and theoretically sound.

You also spent two months in Papua New Guinea for field work. What brings an economist to that remote part of the world?

One of the traditional themes in behavioral economics: social behavior. I worked on a project with Andreas Pondorfer and Susann Adloff in the province of Bougainville. We elicited beliefs about what is socially appropriate behavior in different communities and investigated how this perception and the structure of the social network shape behavior and social-image effects.

While overseas, I also experienced personally how crucial perception is. The Bougainvilleans literally live in a different world. As an outsider, foreign to their fascinating culture and norms, I probably misunderstood many of their gestures and statements – and vice versa. On one day, two lunchs were prepared for me at different places – without me knowing about this. And the lunch that I arranged with my host was surprisingly cancelled. I still don’t fully understand what happened.

briq Beliefs Workshop

Conventional economics assumes that people desire information only because it helps them to make better decisions, and that they adjust their beliefs in response to new evidence. In reality, people often cherish their clearly misguided beliefs, defend them from perceived challenges, and can be hostile to those who hold different views. Examples include climate change denial and the anti-vaccination movement as well as managers who are overconfident about their skills and achievements.

To explain these phenomena, behavioral economists have developed the concept of “belief-based utility”, whereby the utility people derive from their convictions as such can make them cling to wrong beliefs even against all common sense. In the briq Beliefs Workshop, researchers presented and discussed novel findings spanning various facets of this topic – ranging from the role of political convictions to information avoidance and selective memory.

Voting decisions in the U.S. election reflected the match between voters’ and candidates’ moral values.

Several presentations highlighted the importance of beliefs and values as drivers of economic decision-making. Benjamin Enke linked voting behavior in recent U.S. presidential elections to beliefs about what is “right” and “wrong”: Whereas some individuals’ notion of morality is centered around “universal” values such as individual rights, justice and fairness, others emphasize “communal” values like loyalty, respect and tradition. For instance, while Donald Trump’s speeches focused on communal values, Hillary Clinton had rather drawn on universal vocabulary. To a significant extent, voting decisions in the election reflected the match between voters’ and candidates’ moral values, which accounted for 20% of the variation in voting behavior. Enke resumed that recent shifts towards communal values especially in rural areas of the U.S. might contribute to increasing political polarization.

Beliefs and educational choices

With regard to educational choices, Teodora Boneva presented the results of a large-scale survey on undergraduate students’ beliefs about the advantages and disadvantages of pursuing a postgraduate degree. The study uncovered that students whose parents obtained a university education are generally more optimistic than first-generation students: They expect a better social life and financial situation as well as more parental support during their postgraduate education. Moreover, they anticipate a stronger positive effect of a postgraduate degree on future earnings. These findings are important because they can explain why students from better socioeconomic backgrounds are more likely to enroll in postgraduate programs, despite efforts from universities to attract first-generation students.

Competitiveness is predictive of educational attainment and earnings.

Muriel Niederle analyzed the importance of competitiveness for education and labor market outcomes using a large data panel in the Netherlands. Together with her co-authors, she found that competitiveness is at least as predictive for educational attainment and earnings as risk attitudes and a set of commonly analyzed psychological traits like conscientiousness, extraversion and agreeableness. Her findings also confirm the results of her previous work, which showed that gender differences in competitiveness can explain about one fifth of the gender gap in career choices.   

Beliefs and overconfidence

A second focus of the workshop was on how people’s beliefs are skewed towards what they want to be true. For example, Botond Köszegi presented theoretical work based on the observation that people tend to be overconfident and showed that this may result in prejudices. If individuals are generally too optimistic with regard to their ability, they will perceive their group to be discriminated against because they do not receive as much recognition as they think they should. This can explain the occurrence of two socially problematic phenomena: On the one hand, people tend to have overly positive perceptions of groups they are members of, while other groups they do not belong to are perceived too negatively. On the other hand, individuals from different groups disagree over the degree of discrimination.

Positively biased memory of own performance makes managers persistently overconfident.

In his presentation on the role of biased memory for manager overconfidence, David Huffman addressed the question of how overconfidence can persist despite repeated performance feedback. He and his co-authors analyzed the beliefs of store managers who receive about a fifth of their salary in the form of bonus payments, which depend on their relative performance rank among all stores owned by the company. Managers are ranked and awarded bonus payments quarterly, so they receive repeated feedback about their relative performance. Huffman and his team found that managers were persistently overconfident regarding their future performance. Moreover, those with a more positively distorted recall of their previous rank displayed the highest degree of overoptimism. This result provides new evidence for an important mechanism people use to sustain biased beliefs.

However, people’s beliefs are also shaped by their economic experiences. In the context of the 2015 Israeli elections, Moses Shayo demonstrated that participation in financial markets can affect political beliefs: Voters who were randomly assigned either Palestinian or Israeli stocks prior to the election were six percentage points more likely to vote for parties supportive of the peace process than their counterparts who had not received any assets. More than 30% of the effect can be explained by the observation that being exposed to financial markets reduced participants’ opposition to concessions for peace, and increased their awareness of the economic risks and consequences of conflict.

This is only a small selection of the large number of high-quality papers presented at the workshop.
View the full list of presentations here.

Paola Giuliano on the origins and persistence of gender roles

Paola Giuliano at briq, August 31, 2018

A recent strand of the economics literature tried to understand the origin of cultural norms and beliefs. Different studies have documented the continuity of cultural norms over remarkably long periods of time.

In her first of two lectures she recently gave at briq, Paola Giuliano (UCLA Anderson School of Management) reviewed work on the origin of differences in gender norms, with particular emphasis on differences in agricultural technologies. Look to history and you will find plenty of examples of farming communities where women worked as much as men. But in plough-intensive societies, where cultivation and planting relied on physical strength, men were the primary farmers. Today many of these societies still show the same division of labor.

Paola Giuliano reviewed her work which shows that at present the descendants of societies that traditionally practiced plough agriculture have lower rates of female participation in the workplace, in politics and in entrepreneurial activities, as well as greater prevalence of attitudes favoring gender inequality, and even a higher male-to-female sex ratio in every age group.

In the same lecture she more broadly reviewed the most recent empirical evidence on various historical determinants of contemporary difference in gender roles and how these differences are transmitted from parents to children and therefore persist until today. This included work on female labor force participation, fertility, education, marriage arrangements, competitive attitude and even domestic violence.

Although cultural beliefs can be quite persistent, there are also many examples of dramatic changes. This raises the natural question: When does culture change and does it persist? In particular, what determines a society’s willingness to adopt new customs and beliefs rather than hold on to traditions?

In her second lecture, Paola examined a determinant that has been put forth in the anthropology literature to answer this question: the variability of the environment from one generation to the next. She shows that populations with ancestors who lived in an environment with more stability from one generation to the next not only place a greater importance on maintaining tradition today, but they also exhibit more persistence in their traditions over time.