We present a theoretical framework where an investor who values sustainability decides whether or not to invest in a firm. The firm, sustainable or not, discloses information about how green she operates with the aim to obtain investment. We first show that without verifiable certification, sustainability communication is uninformative in equilibrium: any firm can claim to be sustainable, and no claim is credible. Second, we show that credible climate certificates restore informative communication through a unique cutoff equilibrium: only firms above an endogenous sustainability threshold find it worthwhile to certify. Third, when certification standards tighten over time, some firms exit climate initiatives rather than comply with stricter criteria, rationalizing the empirical pattern observed at CA100+ and SBTi. Whether firms anticipate future tightening shapes both their transformation strategy and the long-term effectiveness of climate initiatives. The fixed cost of certification emerges as a key policy instrument, governing the trade-off between broad initial participation and sustained long-term engagement. Our paper adds to the existing literature by incorporating a dynamic model of green certification with heterogeneous firm types and endogenous transformation decisions.
Status: working paper available upon request.
Two agents jointly decide to take a risk and agree on a fair rule for sharing the payoff. When uncertainty resolves, each evaluates the outcome against the certainty equivalent of their individually preferred lottery. What looks fair ex ante can trigger regret ex post, and who regrets depends on risk attitudes. One can think of individuals regretting an insurance contract after a middling loss, or a household disagreeing over a joint investment after a disappointing return. The paper studies this in a classical Borch risk-sharing framework. Each agent has an endorsement threshold given by the certainty equivalent of a reference lottery. A realized division is endorsed when both payoffs clear both thresholds. The main contribution is a characterization of the regret-free zone and its interaction with transfers. The lower-envelope filter identifies which sharing rules are no-objection implementable on their support. When a rule fails, the threshold sum T separates unavoidable conflict from stabilizable disputes: if realized output exceeds T, the objecting agent's shortfall is the exact transfer needed to restore endorsement. For affine contracts, a fixed transfer component can align both endorsement thresholds with T, eliminating all stabilizable one-sided objections.
Status: work in progress.
It has been shown that most people do not update their belief following Bayes rule. However, in these settings people's emotions often affect their decision making. We want to eliminate that factor and examine how and why people still deviate from the behaviour of a risk-neutral Bayesian behavior. Our experiment is designed to examine direction and causality of this derivation. Different to previous work, our focus lies on the role of higher order information. We offer subjects to learn the precision of a previously observed message. By eliciting their willingness to pay to learn as well as to avoid message precision, and by observing how it is incorporated into their updated belief, we describe a market for second order information. Our paper can be linked to Augenblick et al. (2025), as well as to literature on information avoidance.
Status: working paper available upon request.