In recent years, concerns for potential risks on human health related to the overuse of chemical pesticides have encouraged research of alternatives production methods as integrated pest management (IPM) and organic agriculture. Consumer preferences for these practices or for new product characteristics often have been evaluated using stated preference techniques such as Choice Experiment (CE). Nevertheless, it has been found that in these surveys respondents generally report higher hypothetical than real willingness to pay, providing the existence of the so-called “hypothetical bias”. While the presence of this bias has been widely reported in Contingent Valuation, its investigation in CE is still at the beginning. Moreover, in most of the cases, the comparison between hypothetical and real payments treatments has been performed in laboratory settings, employing within-sample approach and providing an initial endowment of money to respondents. This paper contributes to the current literature by presenting an empirical CE study on apples performed in the field (in supermarkets) comparing a hypothetical and a real payment treatment. The latter is done without providing any initial endowment to respondents but asking them to use their own money, that is to pay out of their own pocket. The focus of the survey is to investigate consumers’ preferences for alternative production systems that employ different mixtures of chemicals, natural substances and beneficial microorganisms providing a progressive healthier and safer product. We moved from a conventional to an organic production, passing through an IPM and an innovative technique that employs biocontrol agents. Other investigated attributes are appearance, origin, climate change mitigation practices and price. Moreover, we asked respondents to state their minimum requirements for the attributes’ levels (cut-offs) and to rank the attributes’ importance. Our split sample CE to evaluate apple preferences includes two treatments (hypothetical and real payment) with 96 respondents each. Data were collected in Trentino Province (Italy) during the fall of 2009 by means of a touch-screen computer-assisted self-interviewing system. The results show that consumers’ behavior is significantly different in hypothetical and real treatments, having some parameters a different effect on the probability of purchase in the two treatments. As expected, the price has more influence on the real purchase decision, while alternative methods and the issues of climate change seem weight more heavily in the hypothetical scenarios. Moreover, the coefficient associated to the alternative method that integrates microorganisms into IMP is not statistically significant in both treatments. Regarding the order of attribute importance, the pairwise comparison between the two treatments (hypothetical vs real) indicates that only for the most important attribute (rank 1) the distribution of preferences is statistically different. In any cases, in both treatments the origin is ranked first by the majority of the respondents. Finally, most of interviewed people stated to have cut-offs values in mind when purchasing apples. Regarding methods of production, however, results show that in a real purchasing situation 42% of respondents do not look at the method of production employed at all (with respect to 28% in the hypothetical setting) and, that among those who stated a minimum requirement, 89% violated them. These findings suggest that consumers, besides preferring organic production among other methods, seem to not give much importance on production attribute at the purchasing stage. Furthermore, they do not yet have in mind a clear frame of the other different production methods and their impact on health related aspects.
