Does self-regulation work? Experimental evidence of the reputational incentives of Self-Regulatory Organizations

Self-regulation (SR) is a common way of enforcing quality in markets (such as banking, financial services and several professions) and in a variety of public and private organizations. We provide experimental evidence of the reputational incentives of self-regulatory organizations (SROs) to publicly disclose versus cover-up fraud in an incomplete information environment. We find that observed behaviour is generally consistent with Bayesian equilibrium when subjects are informed about the relative likelihood of fraud detection by a 'vigilant' versus a 'lax' SRO type. In particular, a fraud disclosure equilibrium is supported when subjects are informed that the 'vigilant' SRO is more likely to detect fraud; otherwise, a cover-up equilibrium is supported. However, when subjects are not informed about the relative likelihood of fraud detection by the SRO types (as expected in real SR situations), no equilibrium is strongly supported. Our results suggest that in practice, the reputation-based incentives for effective SR may be inherently ambiguous and weak.