Field Experiments

Laboratory experiments are often criticized for examining a world that is created artificially and therefore existing far away from a real world. Therefore, economists have started to search for possibilities of performing experiments in the real world, i.e. in the field. Harrison and List (2004) summarize the development of field experiments so far – see the paper here or the “Downloads” section. Authors call for complementarity of both laboratory and field experiment rather than for their mutual incompatibility.

Harrison and List define a field experiment as an experiment where the participants act in their natural environment (as opposed to laboratory experiments), where they use natural (as opposed to abstract) means and commodities, moreover, they may not be aware of being a part of an experiment. Thus, a field experiment can be single or even double blind that is not possible to achieve in laboratory (for obvious reasons). The history of field experiments can be traced to 20s and Ronald Fisher’s (1890 – 1962) use of randomization for a precise estimation of yield from agricultural land – in other words, a field. Nomen est omen. Technically speaking, field experiments are older than laboratory ones. The golden age of field experiments in developed countries are 60s and 70s when several governments around developed world began to cooperate with researchers on vast social experiments intervening to many aspects of human life. The American experience with real life and governmentally financed experiments from 1975 till 1981 can be found in Hausman and Wise (1985), where experiments and their results from many fields are elaborated upon: negative income tax (Frank P. Stafford), electricity prices and time of consumption (Dennis J. Aigner), housing subsidies (Harvey S. Rosen) or health industry (Jeffrey E. Harris). The handbook also discusses important methodological or practical questions in the area of field experiments such as ineffective randomization in social experiments (Jerry A. Hausman a David A. Wise) or experimental metodology in medicine (Frederick Mosteller a Milton C. Weinstein) and it also covers the impal of experiments on real economic policy (Ernst W. Stromsdorfer). Field experiments were one of the sources of nontraditional econometrics. J.J. Heckman and D. McFadden received a Nobel price in Economic science for 200 for development of this novel approach to evaluace of social programs (for details on J. J. Heckman see here (in Czech only) or the „Downsloads“ section). The last decade of the 20th century was characterized by many papers which brought field experiments closer to laboratory ones. Laboratory experiments are often criticized for choosing bachelor university students as subjects. This complaint is if possible overcome by inviting professionals who are working in the area of experimental research. For such experiments Harrison and List coin the term artefactual field experiments. It is fair to say that results of bachelor students as subjects are not any worse than those of professionals (for details see the “Downloads” section).

The method of field experiments or, more precisely, natural field experiments, is nowadays used mostly in the area of development economics. The main promoters of this branch of research are the researchers associated with the Abdul Latif Jameel Poverty Action Lab (J-PAL) at the Massachusetts Institute of Technology (MIT). Duflo and Banerjee (see also “Downsloads” section) summarize the advantages of evaluating social programs by using randomization. This method, if appropriately implemented, allows creating a control (counterfactual) group which can be compared to a group that actually participates in an experiment. Michal Bauer and Julie Chytilová are leading researchers in the area of field experiments in the Czech Republic. A large part of Michal Bauer’s course “Economics of Least Developed Countries” at the Charles University is about field experiments and their use. The newly established course “Advanced Experimental Economics” in cooperation of LEE, UEP, CERGE-EI and IES FSV UK dedicates an entire half of the semester to field experiments (see also “Education – UEP” and “Education – Outside UEP” sections). Typical field experiments are those testing discrimination on labor market or discrimination on housing market (see also “Student Works” section, for instance V. Bartoš: “Discrimination, Information and Cognitive Effects: Evidence from a Field Experiment in the Czech Rental Housing Market”).