Experiments in Economics

Alexander Usvitskiy

© HSE University

Why simulate an auction? How do experiments in economics differ from those in psychology? Assistant Professor Alexander Usvitskiy, Head of the Laboratory for Experimental Economics and Finance, discusses the laboratory's activities and research approaches.

— At what point did you become interested in behavioural economics?

— My interest in behavioural economics and game theory began in the second year of my bachelor's programme at the International College of Economics and Finance (HSE ICEF). By the end of the programme, it had developed into a desire to choose an academic career and earn an academic degree. This ultimately led me to enrol in a PhD programme at Florida State University. At the time of my admission, the Department of Economics was known for its strong focus on behavioural and experimental economics. It had its own laboratory and a dedicated research cluster with a dozen professors and a large group of doctoral students. There was considerable freedom in choosing research topics, co-authors, and members of the dissertation committee. There, I successfully passed the qualifying exams and received an enhanced scholarship for the entire period of study (thanks to ICEF!), which allowed me to focus comfortably on research and attend conferences, including the largest in my field. In particular, I had the opportunity to attend an institute for doctoral students from around the world, organised by Vernon Smith, who holds the Nobel Prize for his contributions to experimental economics—a field closely intertwined with behavioural economics and finance.

— That is the title of your laboratory: the Laboratory of Experimental Economics and Finance. What areas of research do the laboratory staff focus on?

— Today, behavioural and experimental economics has developed into a full-fledged field with a wide range of subfields and topics. Accordingly, the research conducted by our laboratory staff is highly diverse and driven by individual interests. That said, I would highlight three main areas: (1) interactions between humans and AI/LLMs; (2) behavioural finance, including cryptocurrency markets and financial decision-making under uncertainty; and (3) behavioural strategies in environments characterised by competition and information asymmetry.

— Is the laboratory engaged in applied research?

— Yes, certainly. We actively develop this area of the laboratory’s work and conduct research for think tanks, industry clients (including banks), and governmental and non-governmental organisations at both the regional and federal level. For example, we often design customised economic experiments for our clients, collect and analyse data, and present the results in the form of reports and discussions, along with recommendations for making changes to business processes.

— Does your involvement in such projects indicate a growing demand for behavioural economics in Russia overall?

— Absolutely. Demand is increasing for both research and specialists in behavioural economics. In turn, we aim to involve as many interested people as possible—both students and established researchers. In particular, we continuously recruit research assistants and doctoral students, working closely with them to develop their skills and actively involve them in our projects.

— Earlier you mentioned Vernon Smith. Could you briefly outline his contribution to economics?

— Beyond the findings of his own research, Vernon Smith’s key contribution was laying the foundation for the methodology we now know as 'economic experiments.' This includes the core principles and practical 'rules' for designing and conducting experiments. I find the analogy with Christopher Columbus, often used in this context, quite fitting: he was not the first to reach America, but it was after his voyages that America appeared on maps and remained there. Likewise, Vernon Smith was not the first to conduct such studies, but it was after his work that experimental methods became firmly established and developed into a distinct field—experimental economics, which, in turn, goes hand in hand with behavioural economics.

— So, what are economic experiments like?

— In their simplest form, economic experiments involve a group of participants who come to a specially equipped computer laboratory and take part in a computer-based simulation for 60–90 minutes. There are two key principles underlying such experiments. First, participants must not be misled in any way. For example, you cannot tell them they are playing against another person when they are actually interacting with a computer or say they will receive a random number between 1 and 100 when in fact everyone receives the same number. While such techniques are sometimes used in psychology, they are not permitted in economic experiments, as it is essential that participants’ decisions and reactions genuinely reflect the conditions described in the instructions. The second rule is that participants must be provided with monetary incentives. The amount they receive should depend on their decisions; otherwise, it is difficult to claim that their behaviour reflects what they truly consider optimal for themselves. In this way, we place participants in a decision-making environment where their choices ultimately determine their payoff.

— Could you give an example of a laboratory experiment?

— A standard example of an economic experiment is an auction. Most people have a general idea of what that is. In a laboratory setting, it works as follows: a group of participants arrives, sits down at computers, and all interactions at this stage are completely anonymous. Next, each participant is assigned a random number representing the value of the object they will receive if they win the auction. After that, participants are asked to place a bid—the amount they are willing to pay to purchase the object. The bids of all participants are then compared, and the highest bidder wins. Their payoff is equal to the difference between the value they were assigned and the amount they bid. Our expectation is that participants will align their bids with their assigned values and act in their own best interest.
 What is the purpose of this? There are many real-world processes, such as tenders, where auctions are the primary mechanism. When the government sells assets through auctions, it is important to do so in the most efficient way possible. Otherwise, it may raise less revenue than it could, especially when auctions are frequent or involve high-value assets. Testing such auction formats in the laboratory first increases the likelihood of identifying the optimal design, thereby maximising the revenue for the organisers.

— Is this the reason why such experiments are useful beyond basic science?

— Yes, among others. It was precisely to test mechanisms like these that the first economic experiments were conducted. More broadly, economic experiments—whether in the laboratory or in the field—provide a unique opportunity to establish causal relationships and identify the factors that influence human behaviour. This, in turn, makes it possible to develop recommendations for improving real-world business processes. It is therefore no coincidence that research in our field is published not only in leading economics journals but also in top management journals such as Management Science.

— Thank you for the interview. Would you like to add anything?

— Thank you, it was a pleasure. I would like to invite readers to take part in experiments conducted at the Laboratory for Experimental Economics and Finance (LEEF), and to join the Telegram channel run by my colleagues.

A recent FES podcast with Alexander Usvitskiy

Register for experiments here.

Telegram channel