Up to recent times, one of the Kaldorian "stylized facts" (Kaldor 1957) of the economies was the stationarity of labor income share. In the late 20th century - however after many decades of relative stability - the labor income share began to decline in most advanced and emerging economies, falling to record lows (IMF, World Economic Outlook, 2017, ch.3). This paper makes two main contributions. After discussing definition and measurement issues, we address the question of the decline of labor income share in the last quarter of century in the Balkan economies (excluding Greece and Turkey). First, we briefly present a short review of the main explanations offered by the literature, categorizing them in two broad groups: the macro interpretation, based on representative agent (industry); the micro interpretation, which emphasizes the heterogeneity of industry. As widely documented, the fall of labor share is a global phenomenon. In fact, using the Total Economy Database (TED, a database developed by the Groningen Growth and Development Centre, University of Groningen, in the early 1990s and since 2007 produced in partnership with The Conference Board) for the period 1990 through 2016, we document that the fall of labor share in most of the Balkan countries. Second, we test the explanations of the labor share fall by looking at their implications and consequences. For instance, most macro explanations rely on technological progress to conclude that it affects the labor share by lowering the user cost of capital, inducing firms to substitute capital for labor, a conclusion which requires an elasticity of substitution larger than 1. Consequently, we should expect an increase in the capital-labor ratio, in labor productivity, and in total factor productivity. Again using TED, we show that the actual evolution are at odds with these predictions in most of the Balkan economies.