An economic experiment
Experimental economics as we know it today can be dated back more than 70 years, when the American economist Edward Hastings Chamberlin (1948) published his paper The Experimental Imperfect Market. Despite his pioneering efforts, it was not until the 1960s that experimental economics developed more widely, with the work of Austin Hoggatt on oligopoly issues, his student Vernon L. Smith on auctions and the behaviour of economic agents in market competition, and Fouraker and Siegel's 1963 work on two-sided bargaining in a monopoly situation. Further development was relatively slow until the late 1970s and early 1980s, when experimental economics experienced unprecedented interest from economists.
An experiment (trial) is characterized as a method of cognition in which, using controlled and controlled conditions, economists investigate real-world phenomena and aim to test a particular proposition under a chosen hypothesis or theory (Guala, 2005).
Compared to observation, experiment has several positive features:
- It allows to investigate the properties of phenomena or objects even under extreme conditions,
- it allows to investigate the observed phenomenon in its "pure" form,
- provides the possibility of its repetition, replication.
Experiments can be divided into three basic groups according to their focus (Kagel and Roth, 1997):
- Speaking to theorists
These experiments are formulated to test the predictions of theorists and to observe them. The controlled environment allows the results to be interpreted directly with the predicted relationships in the theory, and the experiments thus represent a form of dialogue between theory and practice.
- Searching for facts
In this category of experiments, economists study the effects of various variables about which there is little information in the theoretical literature to date. These experiments are often motivated by previous experiments and aim to isolate the factor or case under study and characterize it in more detail.
- Whispering in the ears of princes
In this case, the link between economics and policy. Experiments are motivated, for example, by questions from regulatory agencies in the context of examining the future effects of adjusting market organisation or introducing regulation. In this case, the experimental setting is characterized as if it were a natural setting and we observe the response of the actors in it. We use the results from the experiment to draw inferences necessary for policymakers' decision-making in the economy.
Laboratory experiments
The micro-economic laboratory environment consists, according to V. L. Smith (1982) consists of an environment and an institutional framework. The economic environment consists of agents (1,..., N) and commodities (1,..., K). The institutional framework consists of a set of rules R (r1,...rn). Economic experiments are characterized by features that ensure the credibility and replicability of the experiment. Since economics is the science of the use of scarce resources, the reward for the participants in the experiment is a very important element for obtaining credible information. Reward is most often paid in the form of a financial reward, which consists of two parts: (1) a show-up fee for participating in the experiment and then, after participating in the experiment, (2) a reward determined by the decisions of the subject(s) or other subjects participating in the experiment. Monetary reward is an important motivating element and creates the illusion of the real world.
Laboratory experiments can take two basic forms, either through computers or paper and pencil. Currently, most experiments are conducted in laboratories through computers.
Randomised trials or randomised control trials
Randomised evaluations or randomised control trials (RCTs) have become increasingly important in public policy in recent years. Although it is a standardised and long-standing methodology in the medical sciences, it is a relatively new method with growing popularity in the field of economic decision research.
It is a research method that is based on a deep knowledge of the mechanisms of behaviour of institutions and especially individuals, of the specific practical problem at hand and of previous studies focusing on a similar type of problem.
Randomization, from the name of the method, results from the random assignment of individuals who are affected by a policy or program into groups that are exposed to particular alternatives of a carefully prepared experimental (test) design. Another important feature of RCTs is that their conclusions are based on a detailed analysis of the data collected. Given that individual issues, policies and local contexts differ greatly from each other.
RCTs are not universal and cannot be replicated like a recipe from a cookbook. They are precisely tailored to minimize overlaps with other components of the programs under study, e.g., implemented by multiple service providers or involving multiple service models.
As mentioned earlier, RCTs are primarily a research method involving a team of members in their implementation,
- a coordinator from a specific institution who will supervise the implementation of the study as well as representatives of the institution who are very familiar with the issue, the relations within the institution of the state administration or local government,
- experienced researchers who are able to analyse the scientific literature, are experienced in scientific research methodology and will propose an experimental design based on their knowledge of behaviour
- representatives of the institution who have access to the data,
- analysts with the knowledge to process the data collected
- possibly other members, taking into account the specific nature of the particular study (e.g. programmers, etc.).
Thus, RCTs are a comprehensive tool, usable to investigate and find solutions to diverse problems in the public sector. In this guide, we provide only a brief introduction to RCTs, a schematic overview of the process of designing, implementing and evaluating RCTs. Given the scope and complexity of the issues, it can serve as an initial introduction to RCTs. For those interested in a deeper understanding of the issues, we provide, for several topics, references to literature that can be used for this purpose.
This material is an output of the project KEGA 017EU-4/2021 entitled “Implementation of experimental economics as innovative method for development of new skills for economic education”.