Humankind has pursued happiness for centuries. Given the significance of happiness to people, this study analyzes predictive associations between socioeconomic and psychosocial variables as independent variables and subjective happiness as the dependent variable. The study utilizes data from the Social Science Japan Data Archive (SSJDA), part of the Institute of Social Science at the University of Tokyo, collected in 2022 during the COVID-19 pandemic. The data were reorganized as binary variables and then analyzed in bivariate analyses and subsequently in multivariate models to assess predictive associations between the independent and dependent variables, i.e., subjective happiness. Binary logistic regression analysis identified a model with the following eight independent variables as optimal of all tested models: (1) marital status, (2) self-identified social class, (3) annual household income, (4) affirming one’s own merits, (5) perception that people are trustworthy, (6) having a trustworthy neighbor, (7) feeling lonely, and (8) self-rated health. The discussion section focuses on loneliness, as it is the only variable among the eight predictors that has a statistically significant negative association with subjective happiness. The complex interplay among subjective happiness, its predictors—in particular, loneliness—and the COVID-19 pandemic is explored.