1: \begin{abstract}

2:   We prove that a ``first-order'' Sequential Quadratic Programming

3:   (SQP) algorithm for equality constrained optimization has local

4:   linear convergence with rate $(1-1/\rcond)^k$, where $\rcond$ is the

5:   condition number of the Riemannian Hessian, and global convergence

6:   with rate $k^{-1/4}$. Our analysis builds on insights from

7:   Riemannian optimization -- we show that the SQP and Riemannian

8:   gradient methods have nearly identical behavior near the constraint

9:   manifold, which could be of broader interest for understanding

10:   constrained optimization.

11: \end{abstract}

12: