The Econometrics of Hope

We rarely think of hope as a determinant of economic outcomes. Yet it has been shown to be a motivator of human behavior in the fields of psychology and anthropology. Recent studies have incorporated the longstanding psychological concept of hope into economics, but have failed to pay sufficient attention to measurement and statistical considerations that are required to operationalize hope within empirical econometric frameworks. This study applies advanced measurement techniques on hope survey data to derive hope measures that are meaningful for economic analysis. Our first objective is to ascertain the efficacy of standard survey measures of hope used commonly in the psychology literature. Our second objective is to assess possible sources of heterogeneity between individuals with different demographic characteristics and access to public services. To meet our objectives, we rely on a set of primary data collected in two rural regions of Tanzania with a wealth of mindset questions (e.g., hope, aspirations, self-efficacy, social cohesion). We address our first objective using Item Response Theory (IRT) methods. We find clear evidence that respondents perceive positive and negative questions differently. The former performs well at discriminating people with low levels of hope, while the latter are useful for identifying individuals with high levels of hope. With regards to our second objective, which we address using Differential Item Functioning (DIF) methods, we find significant evidence of heterogeneity. For example, we found high DIF in the responses of adults and adolescents. For instance, if DIF is ignored, adolescents will be erroneously scored as less hopeful than adults. These findings help us understand psychological differences that underpin subgroup responses and allow us to consider them while making causal analyses.