This dissertation contains two chapters on law and economics and one chapter on international trade. An important but under-researched issue for medical malpractice (med-mal) litigation is how physicians' previous medical malpractice experiences affect their behaviour. Using Florida data on closed med-mal claims, I find that if physicians have prior paid claims, their current litigation is resolved faster and is associated with less cost. Having a prior payout does not significantly predict the likelihood or the amount of the current payout. This suggests that ``learning'' occurs as a result of prior med-mal experience. As a theoretical motivation, I developed a dynamic version of the divergent expectations (DE) litigation model. The model predicts, consistent with the data, that physicians have a more realistic analysis of med-mal litigation if they have prior experience. Many robustness checks are carried out to test the results, including using a fixed effect framework, to which the results are robust. In the second chapter, we investigate the impact of ``duty to settle'' rule in predicting patterns in data. Prior models and studies of settlement ignore the insurer's ``duty to settle'' -- the obligation to settle within policy limits if it would be unreasonable to refuse a within-limits settlement offer. We incorporate the duty to settle into a structural model of settlement of medical malpractice claims, and then estimate the model using maximum likelihood methods applied to a Texas closed claims database. Both the data and our model predict: a mass of cases with a settlement demand by the plaintiff exactly at limits; a smaller but still sizeable mass of cases with settlement exactly at limits; very few above-limits payments by insureds; and when above-limits payments are made, they are often by insurers. The model does a reasonable job in predicting data moments, including fractions of cases settled at limits, settled above limits, and tried. Using the model in counterfactual analysis, we predict: (i) with no duty to settle, more cases will be tried; (ii) with strict insurer liability for not settling within limits, there will be fewer trials and more above limits payments by insurers; and (iii) the duty to settle will rarely cause insurers to pay more than the expected value of claims. The third chapter of this dissertation is on international trade. There is a well established literature on the impact of sovereign debt renegotiation on bilateral trade, including Rose (2005) among others. However, there is no study that disentangles impacts of renegotiation on the intensive and extensive margins, where the former is the trade volume of established bilateral trading relationships and the latter is the number of established relationships. This study employs the UNComTrade dataset and debt renegotiation data from the Paris Club for over 150 countries in order to address the impact of a debt renegotiation on the extensive margin of trade. This paper finds that bilateral trade volume declines following a sovereign debt renegotiation. The result is robust to the use of trade lags as instrumental variables to address endogeneity. Consistent with the trade literature, this study documents a negative impact of a debt renegotiation on the trade value using the Tobit approach in a fixed effect model, to appropriately handle censored data. Interestingly, a comparison between the marginal impacts of a debt renegotiation on the extensive and the intensive margins shows that the former effect has at least the same magnitude as the latter. / text