Aleksandra Kuzior

Silesian University of Technology, Gliwice, Poland

E-mail: aleksandra.kuzior@polsl.pl

ORCID 0000-0001-9764-5320 

Wartime business relocation intentions and adaptive capacity in Ukraine: Evidence from multinomial logit and latent class analysis

Yevheniia Polishchuk

Massachusetts Institute of Technology, Cambridge, United States

E-mail: yevheniia.polishchuk@gmail.com

ORCID 0000-0002-6133-910X

David Zámek

Department of Security Studies, Faculty of Entrepreneurship and Law, Pan-European University, Prague, Czech Republic

david.zamek@peuni.cz

ORCID 0009-0003-9020-0830

Miklós Péter Kőmíves

Faculty of Economics and Business, University of Debrecen, Debrecen, Hungary

komives.peter. miklos@econ.unideb.hu

ORCID 0009-0006-8747-6993

Abstract. Russia’s full-scale invasion has transformed the spatial, infrastructural and financial conditions of Ukrainian business activity, making wartime relocation and adaptive capacity critical issues for economic resilience. This study aims to examine how Ukrainian businesses respond to wartime disruption by identifying the determinants of domestic and international relocation intentions and the latent profiles of business vulnerability and adaptation. The empirical analysis is based on survey data from 606 Ukrainian business representatives. It applies multinomial logit modelling to a complete-case sample of 505 observations, together with latent class analysis using 22 binary indicators. The results show that relocation intentions are heterogeneous: 31.7% of respondents reported no relocation intention, 28.7% had not considered relocation, 21.6% considered relocation within Ukraine and 18.0% considered relocation abroad. Domestic relocation is associated with credit risk (RRR = 1.570), logistics disruption (RRR = 2.277), import decline (RRR = 3.498) and reliance on own funds (RRR = 2.101). Foreign relocation is associated with business severity (RRR = 1.599), import decline (RRR = 4.224) and foreign cooperation, as the absence of foreign cooperation reduces the likelihood of relocation abroad (RRR = 0.343). Latent class analysis identifies two profiles: constrained non-investing firms (39.4%) and active adaptive vulnerable firms (60.6%), which differ significantly in relocation attitudes, χ²(3) = 12.83, p = 0.005, Cramer’s V = 0.146.

Received: July, 2025

1st Revision: March, 2026

Accepted: June, 2026

DOI: 10.14254/2071-789X.2026/19-2/4

JEL Classification: D22, L26, M13, R23

Keywords: wartime, business relocation, adaptive capacity, business vulnerability, multinomial logit, latent class analysis, economic resilience