This paper seeks to formalize the treatment effects of strategic choices under the Rubin Causal Model.
Under the guidance of Jorge Guzman, I did the following:
- Cleaning the original code, and creating an open-source R package to apply the framework developed in the paper.
- Use the package created to execute an empirical application of the framework discussed in the paper.
- Create data-replication scripts to reproduce the results discussed in Section 5 of the paper.
- Proof-read the paper prior to submission.
A manager makes a choice to improve their firm’s performance. Only a choice that is both profitable for their firm and unprofitable to competitors creates a competitive advantage. This paper formalizes these type of `strategic’ choices under the Rubin Causal Model. Three new objects are defined: the strategic treatment effect (the benefit predictable from a firm’s characteristics), the strategic determinant function (a mapping of characteristics to strategic treatment effects), and strategic coherence (the importance of resource interactions in strategic treatment effects). Under unconfoundedness, they can be estimated from high-dimensional observational data. I present an application estimating the benefit from choosing venture capital as early-stage financing versus other forms of capital. For equity outcomes, there is no average treatment effect of early-stage VC, but there is significant strategic heterogeneity: some entrepreneurs can benefit substantially from raising early-stage VC, while others are negatively affected. This heterogeneity is predictable from founder, industry, and location characteristics. The role of strategic coherence in this choice is moderate.