Essays on Malnutrition, Savings and Preferences

Abstract: This thesis consists of an introductory part and four self-contained papers addressing three different topics in economics, namely preferences, savings and malnutrition. In Paper I, an instrumental variable approach was employed to analyse how malnutrition affects children’s psychosocial abilities. The study looked specifically at children’s aspirations, self-esteem and self-efficacy. The children’s level of malnutrition was measured by the most commonly used measurement, namely the height-for-age Z score (HAZ). Young Lives panel data from Andhra Pradesh, India, as well as self-reported drought variables were employed as instruments to isolate the causal path. Consistent with previous evidence, the current study’s first-stage relationship showed a negative and statistically significant association between drought and children’s HAZ. The main result revealed that malnutrition had a negative impact on children’s self-efficacy, whereas it did not affect their aspirations or self-esteem. Furthermore, a robustness check showed gender differences in the effect of malnutrition on children’s psychosocial abilities. Thus, the self-efficacy of Indian boys was negatively affected by malnutrition, while that of girls was not. Because these results examine the consequences of malnutrition on children’s non-cognitive abilities, they contribute to a literature which, so far, has mainly focused on the consequences on their cognitive abilities. Also using the Young Lives data set, Paper II presents an investigation into children’s ability to catch up from earlier malnutrition. A comparison was specifically made between two cohorts of children born in 1994–95 and 2001–02, using their HAZ values when they were 8 (in 2002 and 2009, respectively) and then 12 years old (in 2006 and 2013, respectively). The objective was to study the differences between the two birth cohorts and the two age levels. In addition, the paper looked at determining the differences between various points on the distribution of HAZ values, using the quantile regression method. The result showed that the null hypothesis of a homogeneous catch-up effect along the entire distribution of HAZ could be rejected. Furthermore, the result suggested that less malnourished children had lower catch-up rates, whereas more malnourished children had higher ones. This evidence is also consistent with previous results, namely that, despite the economic growth that India has experienced and the hope that malnutrition would have declined as a result, children in the younger cohort suffer from nutritional deficiencies similar to those experienced by the older cohort. Paper III sets out the replication of a study by Brune et al. (2016), who implemented a randomised control trial to understand Malawian tobacco farmers’ saving behaviour. The purpose of the intervention was to spur farmers to deposit some of their harvest income in order to relax the liquidity constraint at the beginning of the subsequent planting season. The intervention was implemented at the same time as the harvest period so that farmers’ saving behaviour could be studied closely. Offering a savings vehicle resulted in more inputs being purchased ahead of the next planting season, leading to higher yields and higher income from the next harvest. A further, and rather surprising, finding was that the extra money used to purchase inputs at the beginning of the next planting season was not from the post-harvest bank deposit: in fact, farmers withdrew most of the harvest income they had deposited well in advance of the next planting season. In this and other aspects, the results from the replication study support the findings of the original study. Finally, Paper IV describes a study on recreational anglers’ willingness to pay for improvements in the characteristics of fishing sites. The objective of the paper was twofold. The first objective was to explore the transferability of anglers’ responses between different angling sites, while the second objective was to estimate the welfare effects for anglers in respect of improvements in angling site characteristics. A choice experiment was used to capture transferability across the angling sites, and to examine anglers’ preferences for changes in angling conditions. A survey was conducted among anglers visiting two water bodies in northern Sweden. The findings showed that the anglers studied had highly heterogeneous preferences, which implies that policy interventions need to consider the specific characteristics of the angling groups they target.