The Dynamics of Hours Worked and Technology

Abstract

We study the relationship between hours worked and technology during the postwar period in the US. We show that the responses of hours to technological improvements have increased over time, and that the patterns captured by the SVAR are consistent with those obtained from an RBC model with a less than unitary elasticity of substitution between capital and labor. Data supports the hypothesis that the observed changes in the response of hours to a technology shock are attributable to changes in the magnitude of the degree of capital-labor substitution, . We argue that the observed time-variation in can arise from changes in the structural composition of sectors (or factors) in a heterogeneous inputs production function or from biases in technological change.Keywords: Real Business Cycles models, Constant Elasticity of Substitution production function, Hours worked, technology shocks.JEL classifi cation: E32, E37, C53. ResumenEn este artículo estudiamos la relación entre horas trabajadas y progreso técnico durante el período de postguerra en los Estados Unidos. Mostramos que la respuesta de las horas a mejoras tecnológicas ha aumentado a lo largo del tiempo y que los patrones capturados por un modelo SVAR son coherentes con los obtenidos por un modelo de ciclos reales con una elasticidad de sustitución entre capital y trabajo menor que 1. Los datos apoyan la hipótesis de que el cambio observado en la respuesta de las horas a mejoras tecnológicas es debido a cambios en la magnitud de la elasticidad de sustitución $\sigma$. Nuestro argumento es que la variación de $\sigma$ puede haber surgido por cambios en la composición estructural de los sectores de actividad económica o, de modo equivalente, la composición de los factores de producción en una función de producción con inputs heterogéneos, o bien por sesgos en la dirección del cambio técnico.Palabras clave: modelos de ciclos reales, función de producción CES, horas trabajadas, choques tecnológicos.

Isoelastic Elasticity of Substitution Production Functions

“…Supplementary Material. We also note that in contrast to our results, Cantore et al (2017) found σ to increase across the consecutive 30-year windows of the post-war US data, from about 0.2 to about 0.7. Their result, however, concentrates on the short-run variation in the data instead of long-run trends as it is based on estimations of a Real Business Cycle model with CES production and a non-smooth, autoregressive specification of factor-augmenting technical change.…”

The U.S. Labor Income Share and Automation Shocks

“…It is interesting to start from monetary policy disturbances. According to Cantore, Ferroni, and León‐Ledesma (), empirical evidence suggests that the LS is counter‐cyclical in response to such disturbances but theoretical DSGE models predict a pro‐cyclical LS. Based on the impulse response functions (IRFs) displayed in Figure , the LS in the JPT model is indeed pro‐cyclical.…”

“…Moreover, parsing the observed gyrations of the LS over five decades allows the present model to come closer to the data in terms of pinning down the LS response to a plethora of shocks. More specifically, Cantore, Ferroni, and León‐Ledesma () find that the LS is counter‐cyclical in response to monetary policy innovations in the data, whereas the New Keynesian model features a monetary policy transmission that is based on a pro‐cyclical LS. These authors argue that several model modifications fail to alter the latter.…”

“…Cantore et al () argue that the hours' response depends on the production function. Cantore, Ferroni, and León‐Ledesma () find that the hours' response has increased over time.…”

Technology Shocks and Non-stationary Hours in Emerging Countries and DSVAR

“…They showed that the statistical inference in SVAR depends on the treatment of hours worked (first differences versus levels). These findings have renewed the debate on the relative contributions of various shocks to the business cycle as well as led researchers in the SVAR literature to draw discernibly contrasting inferences (Cantore, Ferroni, & Leon-Ledesma, 2017; Cantore, Leon-Ledesma, McAdam, & Willman, 2014; Christiano et al, 2003; Francis & Ramey, 2005b; Gali & Rabanal, 2005).…”