Research projects

Project CEMAPRE internal

TitleMethodological Developments in Bootstrap and Panel Data Regression - 2026
ParticipantsPaulo Parente (Principal Investigator), João Manuel Vieira
SummaryThe 2026 project builds on the methodological and applied research developed in previous periods,
focusing on bootstrap inference, mode regression, and expectile regression in panel data models. The
main objectives are to finalise completed work for journal submission and advance ongoing
methodological research.
Paulo Parente, in collaboration with R. J. Smith, will finalise and resubmit the revised article
extending the bootstrap method proposed in Parente and Smith (2018). This work demonstrates that the
Kernel Block Bootstrap applies broadly to models with moment restrictions and allows testing
additional moment constraints and mixed parametric restrictions using generalized empirical
likelihood probabilities. In 2026, the focus will be on addressing referee comments, strengthening
simulations, and preparing the paper for publication.

João Vieira’s work on mode regression for panel data with fixed and interactive fixed effects,
and on the impact of FDI on growth in 19 emerging economies, is complete. In 2026, he will focus on
revising and formatting these articles to prepare them for submission to international journals.

In addition, Paulo Parente will continue two methodological projects. With J. A. F. Machado and J.
M. C. Santos Silva, he will study expectile regression estimators in panel data with individual and
time effects. Separately, with J. M. C. Santos Silva, he will develop a mode regression estimator
under endogenous sampling.

Overall, the 2026 project combines completion and dissemination of finished work with ongoing
development of advanced econometric methods for panel data models.