In this dissertation, I analyze how past and present shocks alter individuals' risk attitudes, beliefs, information processing, and related behaviors in Vietnam. I study behavioral adaptation following three distinct emotionally intense and economically impactful events: disability, exposure to chemical defoliants during the Vietnam War, and proximity to a natural disaster.

Each chapter contributes to a greater understanding of how individuals respond and adapt to experienced shocks over time. Throughout these three essays, I find that the effects of such experiences persist long-term, often for years, if not decades, leading to significant shifts in risk attitudes, beliefs, and behaviors compared to those who remain unaffected. This lasting influence underscores the enduring effects of negative experiences on beliefs and decision-making. By addressing these sustained effects, policies can better inform consumers, mitigate socioeconomic disparities, build resilience, and enhance support for vulnerable populations.

The first essay, co-authored with Dorothee Buehler, Kristin Kiesel, and Ute Rink, uses panel data from the Thailand Vietnam Socio Economic Panel (TVSEP) from 2007 to 2017. This dataset provides detailed survey information on disability events, income, assets, shocks, coping strategies, subjective beliefs, and risk attitudes. This allows us to understand how households with members living with a disability differ in their susceptibility and responsiveness to different types of shocks than those without a disability and how the culmination of these shocks affects the ability of these households to accumulate wealth over time. This work has greatly sharpened my analytical thinking and helped inform snd refine the preceding chapters of this dissertation.

The livelihoods of households with a disabled member have been vastly underexamined in the existing literature. Using this unique household-level panel data, we investigate differences in the immediate effects of shocks on the income of households with a disabled member (DHs) and without a disabled member (NDHs) and the longer-term welfare of these households. When examining the immediate impact of shocks, we find that DHs are more resilient to each incremental health shock than NDHs, but are more vulnerable when confronted with natural shocks (e.g., natural disasters, livestock diseases). While these immediate effects suggest that policies for DHs should offer assistance to cope with natural shocks, our chronic poverty analysis paints a more nuanced picture. DHs are significantly more likely to have experienced chronic or transitory poverty when faced with additional health shocks than NDHs. Conversely, DHs are no more likely to end up in poverty than NDHs when faced with shocks outside of the health domain. Our analysis suggests that DHs have developed a greater resilience to health shocks, but the greater cumulative number of shocks experienced over time makes them more vulnerable to falling into poverty overall.

In the second essay, co-authored with Kristin Kiesel and Ute Rink, I investigate the extent to which negative and emotionally intense past experiences create motivated beliefs and lead individuals to reason in a motivated fashion. Specifically, we examine how life experiences, including exposure to chemical defoliants deployed during the Vietnam War (e.g., Agent Orange), can lead to motivated beliefs and reasoning in the context of vegetable food safety through a lab-in-the-field retail experiment. We then examine the effectiveness of the VietGAP food labeling program in allowing individuals to update their beliefs about vegetable food safety quality, particularly for those with these past life experiences.

Information provision can act as an effective quality signal preventing market failures. Food labels aim to reduce information asymmetries between consumers and producers, but evidence of their effectiveness is mixed. Past personal experiences can alter beliefs and affect how this information is engaged with, processed, and stored in memory. We implement a lab-in-the-field experiment in Vietnam to study the relative effectiveness of both a certified food safety label and retailer claim on purchasing decisions and belief updating and to evaluate the impact of two domain- and non-domain-specific personal experiences, illness from contaminated vegetables and exposure to chemical defoliants used in the Vietnam War, on food safety beliefs and information processing. We show that subjects exposed to chemical defoliants have significantly lower beliefs in food safety quality, while those who were ill do not hold these beliefs. This suggests that intense out-of-domain experiences can impact belief formation across domains. Despite highly valuing food safety information, we find that individuals exposed to chemical defoliants employ several forms of motivated reasoning to avoid or misremember food safety information tied to these past experiences even at a financial cost. Our results suggest that individuals are motivated to maintain prior beliefs as a means of self-preservation where the information processing costs of engaging with information tied to past experiences outweigh the perceived benefits. However, unbiased, disconfirmatory feedback mitigates the extent to which subjects engage in endogenous memory formation. This indicates a viable pathway to enhance policy design and information dissemination to improve decision-making and circumvent entrenched beliefs.

The final essay returns to the TVSEP data with a linkage to satellite data to examine the impact of a natural disaster on individuals' risk attitudes, subjective beliefs about future shocks, and related behavior change. My co-author Marcos Martinez and I focus on the impact of Typhoon Ketsana in 2009, one of the most devastating storms to hit Southeast Asia in recent times. Our analysis reveals that individuals who were exposed to the typhoon become more risk averse a year after landfall. This effect persists up to four years later. We base our findings on this household-level panel data from Vietnam and a difference-in-differences strategy with a continuous treatment variable that exploits variation in the intensity of the typhoon. We conclude that a standard deviation (SD) increase in excess rainfall during the typhoon leads to a 0.23 SD increase in risk averseness one year after landfall, and a 0.24 SD increase four years after landfall. Moreover, individuals exposed to higher excess rainfall are more likely to believe that storms will not transpire in the following five years or will occur with reduced frequency. This result supports the view that the main observed effects indeed reflect updated risk attitudes, rather than changes in the subjective probability structure assigned to the occurrence of storms. Finally, we show that individuals exposed to the typhoon increase their insurance purchasing in the long term. This article contributes to the literature that empirically documents how negative shocks can alter risk attitudes and helps define how climate-related hazards can induce changes in individuals' attitudes and economic behavior.

This dissertation draws on theories of product differentiation from industrial organization and on agricultural economics, more broadly, to study the effect of information on consumer and producer behavior as it relates to agricultural commodities and food products. Agricultural markets are rife with information asymmetries, as complex production practices and modern supply chains and retail systems can obfuscate information about how and where products are produced and the health implications from consuming them. As a result, downstream actors rarely have perfect information about the food they are purchasing, and regulators often face difficulties tracing foodborne illnesses back to the source. The first essay examines the impact of introducing a government certification and labeling program for transitional organic crops on organic and conventional markets. Premiums garnered by transitional labels may reduce conversion costs and increase conversion, but some existing organic producers worry it may dilute the value of the existing organic label. I construct static and dynamic models of vertical product differentiation to explore how transitional certification and commensurate labeling impacts prices and market shares for conventional and organic versions of a generic commodity, and, in turn, organic conversion rates. I index consumers' valuation of the high-quality commodity (i.e., organic) using a flexible Kumaraswarmy distribution instead of the uniform distribution typically used in these types of models, as consumers valuation of quality may vary by commodity and over time. I then estimate outcomes for the domestic strawberry market by assigning values to the model parameters using data obtained from a survey developed in collaboration with a major berry producer and relevant prior research. Results from the static model indicate that, in the short run, transitional certification and labeling slightly decrease prices for organic strawberries, do not alter conventional prices, and improve or hold constant welfare for all consumers. In the long run, transitional labeling significantly increases the share of acreage devoted to organic strawberries, holding total strawberry acreage constant, and subsequently increases prices for both organic and conventional strawberries. The price increase is because the yield gap between conventional and organic strawberries leads to a reduction in total strawberry output as more acres convert. The increased prices benefit strawberry producers but come at the expense of strawberry consumers. These results suggest that a national transitional certification policy may be a useful tool in increasing domestic commodity production.

Essay 2 analyzes the economic losses resulting from the November, 2018 food safety advisory that warned consumers, retailers, and restaurants not to eat, sell, or serve any romaine lettuce or mixed salads containing romaine due to an outbreak of E. coli infections linked to romaine. Detected outbreaks including the one studied here are usually characterized by uncertainty and lack of information as to the scope of implicated products and regions, which often results in broad advisories and widespread damages. I separately estimate losses to growers, processor/shippers, retailers, and food-service operators, in addition losses to society, during and after the advisory period using wholesale data from a cooperating processor that supplies both food-service and retail outlets combined that with public data on spot-market prices and movement provided by the USDA-Agricultural Marketing Service and retail scanner data from the Nielsen Company. I then use regression analysis to predict what prices and sales would have been in the ``but for'' world absent the E. coli advisory, compare these values to the real-world prices and quantities during the advisory period and aftermath for romaine and its substitutes, and compute economic impacts from price changes and lost sales. Due to the structure of grower contracts, growers were minimally impacted by the advisory, while processors and shippers lost approximately \$52.7 million from price and quantity impacts. Retailers amassed \$25.7 million in losses mostly due to pulling product from their distribution channels and shelves in response to the advisory. Conversely, food-service operators were little impacted. I further estimate that societal losses from the Fall 2018 incident were in the range of \$280 to \$350 million. The third essay evaluates consumers' stated and revealed preference for locally produced food products using both a survey and a market-level labeling experiment. The survey, distributed by two Sacramento-area food co-ops, assessed differences in willingness-to-pay (WTP) and intensity of preference for local produce and processed foods and examined which characteristics consumers associate with a ``Local'' label. For the labeling experiment, I introduced ``Local'' shelf labels to products in five processed-food categories in one of these stores for four weeks. Both stores provided weekly sales data by UPC for four weeks prior to, during, and after the experiment, and I further collected detailed information on product claims, product placement, and promotions for each product. The vast majority of shoppers at both stores expressed a preference for locally produced produce and processed foods and indicated they would be willing to pay a premium for them. These preferences, however, did not carry forward to increased sales for products affixed with a ``Local'' label in the retail experiment. An analysis of the sales data in a triple-triple difference framework fails to detect significant average treatment effects. Significant heterogeneous treatment effects suggest that the experimental labels primarily draw attention to or highlight products at the point of purchase, and might be more effective for products that do not already capture consumers' attention through manufacturer's claims, retailer discounts, or prime shelf-positioning.

Taken together, the research in this dissertation provides valuable insights on when the provision of information might help advance policy goals and when it might fall short in that regard. Further, it illustrates how imprecise and overly broad information, like that provided many initial food safety advisories, can generate far-reaching economic losses.

With obesity rates at epidemic levels in the United States, some public health advocates and policy makers have turned to soda taxes as a way to curb soda consumption. A number of U.S. cities, motivated by health-related or fiscal considerations, have implemented or are considering soda taxes that vary by enactment process, scope, and magnitude. These excise taxes are imposed on sugar-sweetened sodas and non-carbonated drinks such as sport drinks, fruit drinks, and iced teas. In some jurisdictions beverages sweetened with artificial sweeteners are also taxed. This study examines how current soda taxes affect the price and volume sales of soda products and related products. In particular, this study explores not only the short-run effects, but also the long-run effects, an important aspect missing in the literature. I also explore the heterogeneity in the impact across product category and package size, and how responses vary among retailers based on their types. Regressivity is a hotly debated topic surrounding an excise tax. This study investigates whether the tax burdens fall more heavily on the relatively poor and ethnic diverse neighborhoods. I additionally explore whether geographic proximity to neighboring untaxed stores alters the pricing behavior of retailers, as well as their soda sales. Finally, I compare the results among current tax policies and further evaluate their effectiveness. In the empirical work, I mainly use a large national IRI scanner data and employ both a Difference-in-Differences approach and a Triple-Difference approach to estimate the effects of soda taxes. Results suggest that soda taxes in all taxed jurisdictions had a significantly positive impact on the price of taxed beverages, with pass-through rates less than 100%, and retailers raised prices gradually over time, rather than pass on the tax all at once. These results indicate that the estimates of the impact of soda taxes based on short-term data might be under stated. The pass-through rates of soda taxes that were passed through a council budget vote are larger than those of taxes that were passed through public referendum. This result is consistent with the hypothesis that the massive media campaigns before an election can draw consumers’ attention to the tax and to avoid losing sales, retailers choose not to raise the price significantly in the immediate aftermath. Moreover, the volume sales of soda drinks in Oakland, San Francisco, and Boulder did not decline due to soda taxes, and Berkeley’s soda sales did not drop until the fourth year. The main goal of soda taxes enacted by these jurisdictions is to curb soda consumption, and the insignificant consumption effects suggest that establishing soda taxes by public vote may not succeed in achieving the goal of improving public health. Results also suggest that consumers did not switch to untaxed and healthier beverages in response to soda taxes; they may drive across the city border and purchase sodas in neighboring untaxed stores. In Philadelphia, there is evidence that the sales reduction of taxable beverages was larger in stores near the city border than in those farther away from the border. Given the possibility that the nearby stores of a taxed city can be indirectly affected, future research on soda taxes should be cautious about using them as controls in the empirical analysis. The impact of soda taxes exhibited huge heterogeneity across store characteristics and product characteristics. Results show that large retailers passed on a smaller proportion of the tax to consumers than small retailers as they may risk losing all profits from consumers’ shopping baskets if they raise the price by a large percentage and cause consumers to shop elsewhere, e.g., in untaxed jurisdictions. The pattern of the heterogeneous impact across products also varies by store type. For example, the pass-through rate for large bottles of sodas was greater than that for small bottles at large retailers, while small retailers saw the opposite result. These findings suggest that different types of stores adopt different pricing strategies and consider cross-product relationships when responding to the soda tax. Therefore, it is necessary for future research to analyze at a disaggregated level, or at least distinguish between store types. Soda taxes might be regressive in relatively large jurisdictions and the results for soda sales suggest that the soda taxes might be less effective at reducing soda consumption among low-income households who tend to purchase more sodas and are more adversely affected by obesity. The tax burdens are also shown to fall more heavily on racially diverse neighborhoods. This study provides insights on how retailers’ responses to soda taxes can vary across different dimensions, and how soda taxes of varying scope and objective can differ in terms of the effects and effectiveness. With more jurisdictions considering an introduction of soda tax, the findings of this study provide important implications for future tax design.

This paper investigates consumer reactions to changes in information provision regarding organic production. Quantitative analyses focus on the actual implementation of mandatory labeling guidelines under the National Organic Program. The unique nature of the fluid milk market in combination with these regulatory changes allows us to place a value on information sets under different labeling regimes. Hedonic price functions provide an initial reference point for analyses of individual responses. A random utility discrete choice model serves as the primary econometric specification and allows consideration of consumer preference heterogeneity along observable household demographics. Our results indicate that the USDA organic seal increases the probability of purchasing organic milk. An initial hedonic price function approach, as well as simulations within the discrete choice framework, suggests that consumers value the change in labeling regulations with regard to organic production. Our results further suggest that consumers substitute away from milk carrying the rBGH-free label. This may indicate that consumers pay less attention to these labels in the time period investigated compared to results found in studies that use earlier time periods.

This paper investigates whether information costs prevent consumers from making healthier food choices under currently regulated nutritional labels in a market-level experiment. Implemented nutritional shelf labels reduce information costs by either repeating information available on the Nutritional Facts Panel, or providing information in a new format. We analyze microwave popcorn purchases using weekly store-level scanner data from both treatment and control stores in a difference-in-differences and synthetic control method approach. Our results suggest that information costs affect consumer purchase decisions. In particular, no trans fat labels significantly increase sales, even though this information is already available on the package. Low calorie labels significantly increase sales, while correlated low fat labels significantly decrease sales, suggesting that labeling response may also be influenced by consumers' taste perceptions. Finally, combining multiple claims in a single label reduces the effectiveness of the implemented labels. Our results provide direct implications for changes to the format and content of nutritional labeling currently considered by the Food and Drug Administration.

Farm Subsidies and Obesity in the United States.

The USDA Organic Seal: What Is It Worth to the Consumer?

Market is Strong for California Processing Tomatoes.