Acta Univ. Agric. Silvic. Mendelianae Brun. 2016, 64, 2087-2094

https://doi.org/10.11118/actaun201664062087
Published online 2016-12-21

Multilevel Modeling in Exploring Institutional Effects on Performance

Vojtěch Olbrecht

Faculty of Business and Economics, Mendel university in Brno, Zemědělská 1, 602 00 Brno, Czech Republic

The article engages in exploring the differences between standard panel regression model and multilevel panel model while estimating growth models on a sample of European food manufacturing companies for period 2004–2013. The article provides an overview of both approaches, critically identifies their strengths and weaknesses and concludes by recommendation on specifics of usage of each of those. Results of models suggest positive impact of political and legal framework and openness of economy and negative impact of government expenditures and economic conditions on firm performance. Comparison of approaches shows that similar results are obtained, though with relative higher significance of multilevel model’s coefficient. Theoretical background of Growth Theory and statistical methods suggest preference of multilevel model with a constrain of data availability.

References

40 live references