Source: Dissertation Abstracts International, Volume: 76-11(E), Section: A.

Adviser: Giuseppe Moscarini.

Access restricted by licensing agreement.

This item is not available from ProQuest Dissertations & Theses.

My dissertation studies household consumption behavior and financial choices. The first chapter studies the efficient distribution of resources in an economy where households face frictions in adjusting the stock of durable goods. When adjustment is costly, a consumer upgrades her durable stock only if her economic resources are high enough. As a result, consumers are risk loving over large risks. In this setting, the efficient allocation is a combination of lotteries and insurance. Society chooses to insure small risks, but to endow every agent of a small probability of a large change in consumption. This theoretical result can be used to understand the popularity of gambling-like products and empirical violations of the full insurance benchmark. I build a theoretical consumption model with adjustment costs and fully characterize the optimal lottery. The optimal degree of inequality depends on how costly is adjustment, the composition of aggregate wealth, and complementarities between durables and remaining consumption goods. Also, I show that the presence of adjustment frictions affects welfare analysis. Welfare gains from reaching efficiency may be decreasing in the dispersion of wealth, and the wealth distribution is a poor proxy for the consumption distribution.

The second chapter studies the dynamic behavior of a household when adjustment of durable goods is costly. In presence of adjustment costs, risk averse consumers facing borrowing restrictions cannot borrow to upgrade durables and may choose to play lotteries. I use this intuition to argue that gambling may substitute for poor credit markets. Moreover, a constrained agent will increase non-durable compared to durable consumption, increasing the durable stock only after large income changes. As a result, non-durable consumption displays excess sensitivity to small shocks and excess smoothness to large shocks. Therefore the model generates non-linearity of consumption in response to shocks, an empirical fact that the standard model cannot explain. Finally, I show that total spending is more reactive to large income changes and discuss implications for consumption policy.

The third chapter (co-authored with Leonardo Gambacorta, Luigi Guiso and Paolo Mistrulli) proposes an empirical test of the presence of biased financial advice when households choose between fixed rate mortgages (FRM) and adjustable rate mortgages (ARM). If households are wary, the relative cost of FRM versus ARM should be a sufficient statistic for a household mortgage choice. The identity of the bank originating the loan should play no role, even if banks differ in the relative cost of originating the two type of mortgages. If households rely on advice to guide their choice, banks may be tempted to bias advice in a direction that is most convenient to them. The relative cost of the two mortgages is no longer a sufficient statistic: characteristics of the bank matter as they proxy for its relative advantage in mortgage origination. We use this implication to test for the presence of biased advice using a unique dataset of 2.4 million mortgages originated in Italy between 2004 and 2010. The choice between FRM and ARM is systematically affected by banks' characteristics, especially when they cannot modify the relative mortgage price at origination. This supports the view that banks are able to affect household mortgage choices both through a price and an advice channel.