Firms choose strategies based on their attributes and industry conditi
ons; therefore, strategy choice is endogenous and self-selected. Empir
ical models that do not account for this and regress performance measu
res on strategy choice variables are potentially misspecified and thei
r conclusions incorrect. I highlight how self-selection on hard-to-mea
sure or unobservable characteristics can bias strategy performance est
imates and recommend an econometric technique that has been developed
to account for this effect. Although this concern applies to a wide ra
nge of strategy questions, to demonstrate its effect I empirically exa
mine if entry mode choice (acquisition versus greenfield) influences f
oreign direct investment survival. In specifications that do not accou
nt for self-selection, I find that greenfield entries have survival ad
vantages compared to acquisitions. This confirms previous findings. Ho
wever, the significance of this effect disappears once I account for s
elf-selection of entry mode in the empirical estimates. The results co
nfirm that estimates from models that do not account for self-selectio
n of strategy choice can lead to incorrect or misleading conclusions.