Ayer explicaba en clase el modelo de Heckscher-Ohlin y cómo el comercio es uno de los causantes del aumento de la desigualdad entre los salarios de los trabajadores cualificados y los no cualificados, y que también tiene su papel en la desigualdad entre trabajo y capital. Pero les señalaba que la evidencia empírica dice que el principal responsable es el cambio tecnológico.
Hace pocos años, Bergholt, Furlanetto y Maffei indicaban quiénes eran los culpables (al menos en este siglo XXI): "most of the pre-crisis decline can be attributed to automation, while firms’ rising market power has been the main source of lower labor shares since the Great Recession."
Hoy he visto que la Organización Internacional publicaba un boletín sobre los últimos datos acerca de esta cuestión. En concreto indica:
The labour income share (the proportion of total income earned by workers in an economy) has experienced a decline over the past two decades of 1.6 percentage points. New estimates indicate that the labour income share continued to decrease since 2019, declining to 52.3 per cent in 2022 and remaining at that level in 2023 and 2024. This is 0.6 percentage points below the prepandemic level. This decline, while modest in percentage point terms, represents a sizeable and persistent shortfall in labour income (...)
Analysing the impact of technological innovations over the last two decades across countries with the required data, we find that while the innovations have produced persistent increases in labour productivity and output, they can also reduce the labour income share. This is consistent with automation-based technological innovations driving the aggregate effects. Hence, if these historical and economic patterns were to persist, absent a stronger policy response across a wide range of relevant domains, the recent breakthroughs in generative AI could exert further downward pressure on the labour income share. Nonetheless, the results presented should not be taken as a prediction. First, there is uncertainty about the type and size of shock that AI will represent, which could be very different than what has been observed in the recent past. Second, the process of technological innovation can be steered and influenced through policies that mitigate potential adverse impacts on inequality to ensure that the benefits of technological progress are widely distributed.
Algunas gráficas que aportan (haz clic sobre ellos para agrandarlos):