This article introduces conjoint analysis as a method for causal inference in survey experiments, particularly useful for understanding multidimensional choices in political science. Conjoint analysis allows researchers to estimate the causal effects of multiple treatment components simultaneously, addressing the limitation of classical survey experiments that cannot identify which components of a multidimensional treatment are influential. The authors propose a new causal estimand, the Average Marginal Component Effect (AMCE), which can be nonparametrically identified and estimated from conjoint data using a fully randomized design. They also develop diagnostic checks to validate the identification assumptions. The article applies these techniques to empirical examples on voter preferences and attitudes toward immigrants, demonstrating the value of conjoint analysis in political science research.This article introduces conjoint analysis as a method for causal inference in survey experiments, particularly useful for understanding multidimensional choices in political science. Conjoint analysis allows researchers to estimate the causal effects of multiple treatment components simultaneously, addressing the limitation of classical survey experiments that cannot identify which components of a multidimensional treatment are influential. The authors propose a new causal estimand, the Average Marginal Component Effect (AMCE), which can be nonparametrically identified and estimated from conjoint data using a fully randomized design. They also develop diagnostic checks to validate the identification assumptions. The article applies these techniques to empirical examples on voter preferences and attitudes toward immigrants, demonstrating the value of conjoint analysis in political science research.