Chapter 1 is a critique of a highly influential paper in corporate governance. Chhaochharia and Grinstein (2009) estimate that CEO pay decreases 17% more in firms that were not compliant with the recent NYSE/Nasdaq board independence requirement than in firms that were compliant.We document that 74% of this magnitude is attributable to two outliers out of 865 sample firms.In addition, we find that the compensation committee independence requirement increases CEO total pay, particularly in the presence of effective shareholder monitoring.Our evidence casts doubt on the effectiveness of independent directors in constraining CEO pay as suggested by the managerial power hypothesis.In chapter 2, we investigate whether the earnings overstatements that led to the Sarbanes-Oxley Act (SOX) may have been an intended consequence of pay-for-performance. We find that incentives were higher when current shareholders stood to benefit from overstatements by selling their shares at inflated prices. Incentives also fell in response to the additional costs imposed by SOX, and the decrease is concentrated in firms whose shareholders benefit from overstatements. If overstatements were a symptom of the agency conflict, incentives should have increased around SOX to induce more productive effort as managers voluntarily cut back on overstatements.The empirical evidence thus rejects the view that earnings overstatements prior to 2002 were an unintended consequence of pay-for-performance.In chapter 3, we explore the relationship between obesity and household credit risk. Obesity is a known health risk factor and carries a social stigma. Its presence provides a potentially informative signal about individuals;; choices and preferences. Using NLSY survey data, we estimate that the loan delinquency rate among the obese is 20% higher than among the non-obese after controlling for numerous observable, prohibited, and - to lenders - unobservable credit risk factors.The economic significance of obesity for delinquencies is comparable to that of job displacements. Obesity is particularly informative about future delinquencies among those with low credit risk. In terms of channels, we find that the obesity effect is at least partially mediated through poor health, but is not attributable to individuals;; time preferences.