The welfare concept is central to economics and to some topical problems, requiring interdisciplinary research. Psychology, biology and pedagogics may be important contributors. Economists have overemphasized economic, and among these consumption, aspects. A learning process to reduce unhealthy consumption remains topical. On the production side the diversity of labor types has been neglected. Here a central problem is job choice by maximizing welfare. Variables needed are capabilities required by jobs and personality traits. We ignore the number of independent characteristics. Path analysis with inherited and learnable traits may be extended by data on grandparents and on non-cognitive capabilities. A more precise production function of education is badly lacking. Two methods of measuring welfare functions and the shape of the latter are discussed.