Anastasia Papadopoulou

We study an investment-trading game in which entrepreneurs decide whether to acquire private information before investing in a technology with uncertain profitability. Entrepreneurs may be forced by a liquidity shock to sell their capital to financial markets before the fundamental is revealed, while traders price capital using public information and aggregate investment. We show that entrepreneurs acquire private information only when the probability of the liquidity shock is below a threshold, which is increasing in the precision of traders’ public signal. Relative to a benchmark economy in which traders are fully informed, entrepreneurs inefficiently underinvest in private information. Private information acquisition improves the informational content of asset prices, but weakens the conditional efficiency of entrepreneurs’ investment decisions. Finally, better informed traders increase entrepreneurs’ incentives to acquire private information.

The Klee–Kandinsky method is one of the most widely used procedures for inducing minimal group identities in experiments. It is typically treated as a simple way to create arbitrary but behaviourally meaningful social identities. This paper explores the reliability and robustness of this procedure using a meta-analysis of prior experiments and a preregistered online survey experiment with 1,136 participants. Study 1 yields two main findings. First, implementation and reporting practices vary substantially across the literature. Second, estimates show that the Kandinsky group is significantly larger than the Klee group, with the magnitude of the imbalance varying across regions. Study 2 tests the procedure under standardized conditions. Participants were again significantly more likely to be assigned to Kandinsky than to Klee, and assignment was associated with several observable characteristics, including risk tolerance, region, ethnicity, age, and familiarity with Kandinsky. These findings suggest that the Klee–Kandinsky procedure should not be assumed to generate uniformly balanced or fully arbitrary groups. The paper shows how routine design choices and incomplete reporting can affect the reliability of experimental evidence.

People generally sign an employment contract when joining an organization. At the same time, they often enter into a social contract—an implicit agreement to go beyond the formal job description to support the organization when needed. Such behaviors, known as organizational citizenship behaviors (OCBs), include volunteering for non-contractual tasks that are essential for institutional functioning but not formally assigned. We investigate two factors that influence individuals’ willingness to engage in discretionary OCBs: (i) perceived fairness in compensation, and (ii) the observability of voluntary effort opportunities. We implement a pre-registered, real effort gift-exchange experiment with six treatment conditions that vary systematically when principals and agents learn about a voluntary effort task. We find strong support for reciprocity based predictions at both the extensive margin (whether agents provide voluntary effort) and the intensive margin (how much effort they provide). Fairness perceptions have heterogeneous, treatment- dependent effects. Notably, agents frequently provide voluntary effort even when principals cannot observe it, pointing to intrinsic motivation beyond purely strategic reciprocity. Information structure matters: the proportion of non-providers is highest in the treatment where principals receive information only immediately before the splitting decision.