The Structural Bite: A Methodological Framework for Minimum Wage Studies using Spanish Administrative Data

Abstract

We study the employment effects of the 22% increase in the Spanish minimum wage in 2019, focusing on young workers. Using census-grade administrative tax data covering the universe of formal wage bills and employment (Models 190/390 linked to personal income tax records), we construct several measures of treatment intensity, including two structurally grounded bite indicators based on the incidence of young minimum-wage workers and the implied increase in the wage bill obtained via Exponential Tilting. Difference-in-differences estimates with two-way fixed effects, dynamic event-study specifications, and robust confidence intervals from the HonestDiD framework all point to the same conclusion: the reform did not generate net disemployment effects for young workers. Point estimates of the elasticity are small and often positive, and confidence internals comfortably include zero even with sizable deviations from parallel trends. A triple-difference design exploiting pre-existing tourism dependence further shows that the sharp employment collapse of 2020 is primarily explained by the COVID-19 shock operating through tourism-intensive sectors, rather than by the minimum-wage hike itself. Our results suggest that, in the macroeconomic and institutional environment prevailing in Spain in 2019, with the minimum wage rising to around 60% of the average wage in a recovering economy, the labour market absorbed a large discrete increase in the wage floor without destroying aggregate youth employment. More broadly, the paper highlights how the choice of treatment definition, the use of census-grade data, robust DiD inference, and explicit modelling of concurrent shocks can shape conclusions about the effects of minimum-wage policies.

Figures (4)

• Figure 1: Distributional Properties and Identification Power of Treatment Intensity Measures.Left Panel: Violin plots displaying the density and individual distribution (jittered points) of the four treatment intensity measures across 90 region-sector cells in 2018. Note the high concentration of values near 1.0 for Youth Incidence (Panel A), indicating high saturation, versus the long right tail of the Monetary Gap (Panel C). Right Panel: Coefficient of Variation (CV) for each measure. The Monetary Gap ($CV=1.01$) exhibits the highest relative variance, providing the strongest identification signal. In contrast, Youth Incidence ($CV=0.12$) falls below the 0.15 threshold (red dashed line), flagging a potential risk of low identifying variation.
• Figure 2: Dynamic Impact of Minimum Wage Exposure on Youth Employment (Event Study Estimates).
• Figure 3: Dynamic Effects of Minimum Wage Exposure on Youth Employment (Event Studies across Treatment Definitions). Each panel plots the estimated coefficients ($\beta_k$) and 95% confidence intervals from the dynamic specification: $\ln(Y_{it}) = \sum_{k \ne 2018} \beta_k (D_i \times \mathbb{1}_{t=k}) + \alpha_i + \lambda_t + \varepsilon_{it}$. The reference year is 2018. Panel A (Youth Incidence) and Panel C (Monetary Gap) show the structural measures; note the significant positive pre-trends (2014--2017) reflecting cyclical recovery in high-exposure sectors. Panel B (Kaitz Index) and Panel D (Sectoral Incidence) display the institutional proxies. The vertical dashed line marks the introduction of the 22% minimum-wage hike in 2019. Post-treatment coefficients do not exhibit a sharp structural break into negative territory, supporting the null of aggregate disemployment.
• Figure 4: Sensitivity of Employment Estimates to Violations of Parallel Trends (HonestDiD). Each panel plots the robust 95% confidence intervals for the treatment effect ($\beta_{2019}$) as the allowance for parallel trend violations ($\bar{M}$) increases. $\bar{M}=0$ assumes exact parallel trends (standard DiD). Higher values of $\bar{M}$ allow the counterfactual trend difference to be up to $\bar{M}$ times the maximum observed pre-trend. The red dashed line marks the null effect. Note that for the structural measures (Panels A and C), the confidence intervals include zero even under strict assumptions ($\bar{M}=0$), indicating a robust null result. In contrast, the institutional proxies (Panels B and D) show a significant positive effect at $\bar{M}=0$ that vanishes (crosses zero) with minimal deviations ($\bar{M}=0.5$), revealing the fragility of those findings.