A multi-period perspective on tourism’s economic contribution – a regional input-output analysis for Sweden

Abstract

Purpose Previous studies on tourism input-output (IO) primarily focus on a single year’s snapshot or utilize outdated IO coefficients. The purpose of this paper is to analyze the multi-period development of regional tourism capacities and its influence on the magnitude of the industry’s regional economic contribution. The paper highlights the importance of applying up-to-date IO coefficients to avoid estimation bias typically found in previous studies on tourism’s economic contribution. Design/methodology/approach For the period 2008-2014, national IO tables are regionalized to estimate direct and indirect economic effects for output, employment, income and other value-added deffects. A comparison of Leontief inverse matrices is conducted to quantify estimation bias when using outdated models for analyzing tourism’s economic contribution. Findings On the one hand, economic linkages strengthened, especially for labour-intensive sectors. On the other hand, sectoral recessions in 2012 and 2014 led to an economy-wide decline of indirect effects, although tourists’ consumption was still increasing. Finally, estimation bias observed after applying an outdated IO model is quantified by approximately US$4.1m output, 986 jobs full-time equivalents, US$24.8m income and US$14.8m other value-added effects. Research limitations/implications Prevailing assumptions on IO modelling and regionalization techniques aim for more precise survey-based approaches and computable general equilibrium models to incorporate net changes in economic output. Results should be cross-validated by means of qualitative interviews with industry representatives. Practical implications Additional costs for generating IO tables on an annual base clearly pay off when considering the improved accuracy of estimates on tourism’s economic contribution. Originality/value This study shows that tourism IO studies should apply up-to-date IO models when estimating the industry’s economic contribution. It provides evidence that applying outdated models involve the risk of estimation biases, because annual changes of multipliers substantially influence the magnitude of effects.