This study uses a sample of Central American firms to estimate a set of multilevel regressions following the method of Goldszmidt, Brito, and de Vasconcelos (2011) to explore the relation between return on assets and factors relating to host-country, industry, firm, and year. The results indicate that firm accounts for between 45% and 50%, industry between 10% and 17%, and country between 5% and 8% of variance depending on the specification. The results are consistent with McGahan and Porter (1997), who find a large industry effect, and Goldszmidt et al. (2011) and Makino, Isobe, and Chan (2004), who find important location effects. However, findings in this study suggest that previous studies might underestimate the effect of location, since the influence of country is important in this sample of firms operating in small, neighboring, similar nations. Extant research either omits most Central American countries or groups them together under the same category.

, , , , ,
INCAE Top Articles
Journal of Business Research
Incae Business School

Ketelhöhn, N, & Quintanilla, C. (2012). Country Effects on Profitability: A Multilevel Approach Using a Sample of Central American Firms. Journal of Business Research, 65(12), 1767–1772. doi:10.1016/j.jbusres.2011.10.036