This thesis researches the economics of type 2 diabetes in middle-income countries (MICs). Given the high prevalence of type 2 diabetes in MICs, in-depth country specific analysis is key for understanding the economic consequences of type 2 diabetes. The thesis consists of four studies with the unifying theme of improving the understanding of the causal impact of diabetes on economic outcomes. Study (1) provides an updated overview, critically assesses and identifies gaps in the current literature on the economic costs of type 2 diabetes using a systematic review approach; study (2) investigates the effects of self-reported diabetes on employment probabilities in Mexico, using cross-sectional data and making use of a commonly used instrumental variable approach; study (3) revisits and extends these results via the use of a fixed effects panel data analysis, also considering a broader range of outcomes, including wages and working hours. Further, it makes use of cross-sectional biomarker data that allow for the investigation of undiagnosed diabetes. Study (4) researches the effect of a diabetes diagnosis on employment as well as behavioural risk factors in China, using longitudinal data and applying an alternative identification strategy, marginal structural models estimation, while comparing these results with fixed effects estimation results. The thesis identifies a considerable economic burden of diabetes in middle-income countries and uncovers several inequities affecting women, the poor and the uninsured. Biomarker results indicate that the adverse effects are limited to those aware of their diabetes. Finally, women are also found to achieve fewer positive changes of their behavioural risk factors after a diabetes diagnosis than men, offering a potential explanation for their more adverse employment outcomes compared to men. To reduce the economic burden, the groups most affected by the identified inequities should be targeted. Further, the underlying reasons for the found sex differences need to be identified.