We investigate the usefulness of the dice game paradigm to public administration as a standardized way of measuring (dis)honesty among individuals, groups, and societies. Measures of dishonesty are key for the field’s progress in understanding individual, organizational, and societal differences in unethical behavior and corruption. We first describe the dice game paradigm and its advantages and then discuss a range of considerations for how to implement it. Next, we highlight the potential of the dice game paradigm across two diverse studies: prospective public employees in Denmark (n = 441) and prospective public employees in 10 different countries with very different levels of corruption (n = 1,091). In the first study, we show how individual-level behavioral dishonesty is very strongly negatively correlated with public service motivation. In the second study, we find that widely used country-level indicators of corruption are strongly correlated with the average behavioral dishonesty among prospective public employees. The results illustrate the importance of the validated dice game paradigm to shed light on core questions that link micro- and macro-level dynamics of dishonesty and corruption in the public sector.

We study the role of self-selection into public service in sustaining honesty in the public sector. Focusing on the world’s least corrupt country, Denmark, we use a survey experiment to document strong self-selection of more honest individuals into public service. This result differs sharply from existing findings from more corrupt settings. Differences in pro-social versus pecuniary motivation appear central to the observed selection pattern. Dishonest individuals are more pecuniarily motivated and self-select out of public service into higher-paying private sector jobs. Accordingly, we find that increasing public sector wages would attract more dishonest candidates to public service in Denmark. (JEL D73, H83, J31, J45)

Are country-level differences in corruption related to the dishonesty level of individuals entering public service? Recent studies have found that dishonest individuals self-select into public service in high-corruption settings. Little is known, however, about what is driving this pattern and whether a similar pattern exists in low-corruption settings. This paper examines selection into public service in the world's least corrupt country, Denmark. We subject a relevant student population to a standard experimental dishonesty task and develop a novel method to estimate individual-level dishonesty from the experimental data. We then relate estimates of dishonesty to subjects' job preferences and characteristics. In contrast to previous findings, dishonest individuals in low-corruption Denmark are less likely to want to enter public service. This self-selection is not related to risk-aversion or ability. Instead, we find that dishonest individuals who self-select into higher paid private sector careers such as finance are less altruistic and place a higher weight on their own earning opportunities. Accordingly, counterfactual wage questions suggest that higher public sector wages would attract more dishonest candidates to the public sector in Denmark.

Denotes co-first authors.The not-too-distant future may bring more ubiquitous personal computing technologies seamlessly integrated into people's lives, with the potential to augment reality and support human cognition. For such technology to be truly assistive to people, it must be context-aware. Human experience of context is complex, and so the early development of this technology benefits from a collaborative and interdisciplinary approach to researchwhat the authors call "hybrid methodology"-that combines (and challenges) the frameworks, approaches, and methods of machine learning, cognitive science, and anthropology. Hybrid methodology suggests new value ethnography can offer, but also new ways ethnographers should adapt their methodologies, deliverables, and ways of collaborating for impact in this space. This paper outlines a few of the data collection and analysis approaches emerging from hybrid methodology, and learnings about impact and team collaboration, that could be useful for applied ethnographers working on interdisciplinary projects and/or involved in the development of ubiquitous assistive technologies.