arXiv subjects

Zachary Porreca

Publications and source records attributed to Zachary Porreca.

2 recordsLinked to original sources

arXiv2024

Bride Kidnapping and Informal Governance Institutions

Bride kidnapping is a form of forced marriage in which a woman is taken against her will and coerced into accepting marriage with her captor. Post-Soviet Kyrgyzstan has seen a large increase in the prominence of this practice alongside a revitalization of traditional values and culture. As part of this resurgence of Kyrgyz identity and culture, the central government has formalized the authority of councils of elders called aksakals as an arbitrator for local dispute resolution -- guided by informal principles of tradition and cultural norm adherence. Bride kidnapping falls within the domain of aksakal authority. In this study, I leverage data from a nationally representative survey and specify a latent class nested logit model of mens' marriage modality choice to analyze the impacts that aksakal governance has on the decision to kidnap. Based on value assessment questions on the survey, men are assigned to a probability distribution over latent class membership. Utility function parameters for each potential marriage modality are estimated for each latent class of men. Results suggest that living under aksakal governance makes men 9% more likely to obtain a wife through bride capture, with men substituting kidnapping for choice marriage modalities such as elopement and standard love marriages.

econ.GN↗

arXiv2024

A Note on Uncertainty Quantification for Maximum Likelihood Parameters Estimated with Heuristic Based Optimization Algorithms

Gradient-based solvers risk convergence to local optima, leading to incorrect researcher inference. Heuristic-based algorithms are able to ``break free" of these local optima to eventually converge to the true global optimum. However, given that they do not provide the gradient/Hessian needed to approximate the covariance matrix and that the significantly longer computational time they require for convergence likely precludes resampling procedures for inference, researchers often are unable to quantify uncertainty in the estimates they derive with these methods. This note presents a simple and relatively fast two-step procedure to estimate the covariance matrix for parameters estimated with these algorithms. This procedure relies on automatic differentiation, a computational means of calculating derivatives that is popular in machine learning applications. A brief empirical example demonstrates the advantages of this procedure relative to bootstrapping and shows the similarity in standard error estimates between this procedure and that which would normally accompany maximum likelihood estimation with a gradient-based algorithm.

econ.EM↗