Multilevel Modeling for Economists: Why, When and How

NRU Higher School of Economics , 2020. No. 233.

Oshchepkov A. Y., Shirokanova A.

Multilevel modeling (MLM, also known as hierarchical linear modeling, HLM) is a methodological frameworkwidely usedin the social sciences to analyze data with a hierarchical structure,where lower units of aggregation are ‘nested’ in higher units, including longitudinal data.In economics, however, MLM is usedvery rarely.Instead, economists use separate econometric techniques including cluster-robust standard errors and fixedeffects models. In this paper, we reviewthe methodological literature and contrast the econometrictechniques typically used in economics with the analysis ofhierarchical data using MLM. Ourreview suggests that economic techniques are generally lessconvenient, flexible, and efficientcompared to MLM. Theimportant limitation of MLM, however, isits inability to deal with the omitted variable problem at the lowest level of data, while standard economic techniques may be complemented by quasi-experimental methods mitigatingthis problem.It is unlikely, though, that this limitation can explain and justify the rare use of MLM in economics.Overall, we conclude that MLM has been unreasonably ignored in economics,and we encourage economists to apply this frameworkbyproviding ‘when and how’ guidelines.

Keywords: модель с фиксированными эффектами, смешанные модели, многоуровневая модель, hierarchical linear modeling, иерархические линейные модели, multilevel modelling, модель с изменяющимися коэффициентами, fixed effects, random effects, mixed effects

Publication based on the results of: Preferences and non-cognitive components of human capital and the functioning of the Russian labour market (2020)