We conduct a comprehensive analysis of health determinants at the individual and workplace levels. Using a new individual-level German data set, we investigate the influence of these determinants on health, including collegiality, personality traits as measured by the Big Five, commitment to the company and job characteristics, while controlling for a set of standard sociodemographic and employment variables. We are interested in which determinants are important and which are less influential, whether interaction effects should be taken into account and whether the results depend on the modeling and estimation method used. Among the Big Five factors, conscientiousness, agreeableness and emotional stability are positively correlated with good overall health. The influence of job characteristics such as having substantial decision-making authority, not having physically demanding tasks, having pleasant environmental conditions, facing minimal time pressure and not being required to multitask are also positive. If employees frequently receive help when needed from their colleagues and do not feel unfairly criticized by others in the firm, they usually have fewer health problems. Each Big Five item influences mental health, whereas no statistical significance could be found for these items’ relationships with the number of days workers were absent due to sickness, except for neuroticism. These results are, for the most part, robust to different modeling and estimation methods.

我们在个人和工作场所层面对健康决定因素进行全面分析。我们使用新的个人层面的德国数据集，研究了这些决定因素对健康的影响，包括大学风气，由“五巨头”衡量的人格特质，对公司的承诺和工作特征，同时控制了一组标准的社会人口统计学和就业情况变量。我们关注哪些决定因素很重要而哪些影响较小，是否应考虑相互作用的影响以及结果是否取决于所使用的建模和估计方法。在五大因素中，尽责，乐于助人和情绪稳定与总体健康状况正相关。工作特征的影响，例如拥有强大的决策权，没有体力要求的任务，令人愉悦的环境条件，最小的时间压力以及不需要执行多任务也是积极的。如果员工经常在需要时从同事那里获得帮助，而又不觉得公司中的其他人受到不公正的批评，那么他们的健康问题通常会更少。每个大五项都会影响心理健康，而除了神经质外，这些项与因疾病缺勤天数之间的关系没有统计学意义。这些结果在很大程度上对不同的建模和估计方法具有鲁棒性。如果员工经常在需要时从同事那里获得帮助，而又不觉得公司中的其他人受到不公正的批评，那么他们的健康问题通常会更少。每个大五项都会影响心理健康，而除了神经质外，这些项与因疾病缺勤天数之间的关系没有统计学意义。这些结果在很大程度上对不同的建模和估计方法具有鲁棒性。如果员工经常在需要时从同事那里获得帮助，而又不觉得公司中的其他人受到不公正的批评，那么他们的健康问题通常会更少。每个大五项都会影响心理健康，而除了神经质外，这些项与因疾病缺勤天数之间的关系没有统计学意义。这些结果在很大程度上对不同的建模和估计方法具有鲁棒性。