LTReikšminiai žodžiai: Autoregresiniai slenkamųjų vidurkių modeliai; Darbo rinka; Kointegracija; Kointegravimas; Minimalus darbo užmokestis; Slenkstiniai autoregresiniai modeliai; Įvertinimas; Cointegration; Estimation; Labour market; Minimum wage; Threshold autoregressive models.
ENThis paper presents a comprehensive and non-standard labour market analysis based on univariate and multivariate models for wages. The novelty of this paper lies in the use of non-normalized cointegrating vectors for labour market analysis. Wages are the basis of labour market models, as well as the key factor for employees and employers; therefore, the central analytical axis is a classical wage bargaining process, where one side requires and the other side proposes a certain level of wages. Analysis is divided into two parts: foremost, a careful analysis of Lithuanian wages is conducted and a univariate model for the investigation of interactions between the minimum wage and the rest of the wages is proposed; only after the minimum wage model is drafted can the multivarate model for the whole economy be built up. Briefly, the methodology used in this article can be annotated as a synthesis of sequential theoretical and empirical considerations that combine the results of theoretical macroeconomics with data-generating patterns and stylized facts. The model is considered as the final one only if macro-theory preconditions, statistical prerequisites, and stylized real-world requirements are met and fulfilled. In addition, this article gives an example and a quantitatively, as well as qualitatively, motivated suggestion as to how to incorporate minimum wages into econometric models and puts forward an explanation for why it is necessary to include minimum wage dynamics into labour market analysis. The article is nothing but an empirical case study that demonstrates how many minor details have to be taken into account until a realistic labour market model is built up. Although the paper deals with the labour market, the suitable application of time series methods is the main subject of the analysis. [From the publication]