Analysis you can check.
Three studies of European labour data and one demonstration of a survey method. None of them is client or employer work. Each states its question, sources, transformations and limits, and each links to the data and the code.
The four pieces
- 01 · Labour markets
Europe’s hiring squeeze did not end. It moved south.
Unemployment and job-vacancy rates, 26 EU countries, 2010–2025. Beveridge curves as connected scatterplots; a tightness index; who moved which way after 2019.
Eurostat une_rt_a · jvs_a_rate_r2 - 02 · Labour mobility
One in six working-age Romanians lives elsewhere in Europe.
Working-age citizens living in another EU or EFTA country, 2010–2024, as a share of all such citizens, and their employment rates against those of nationals at home.
Eurostat lfst_lmbpcita · lfst_lmbercita · lfsa_pganws · lfsa_ergan - 03 · Cost of employment
The price of an hour of work is converging.
Hourly labour costs in the EU-27, 2008–2025: levels on a log scale, growth against starting point, and what the split between wage and non-wage costs does and does not mean.
Eurostat lc_lci_lev - 04 · Research methods
What a pooled model hides.
Choice-based conjoint on 600 simulated respondents: multinomial logit, a two-class latent-class model fitted by EM, willingness to pay and an interactive share simulator.
Synthetic data · clearly labelled
What these pieces have in common
- Sources are named to the table. Every figure gives the Eurostat dataset code, the breakdown used, the period and the retrieval date.
- Missing stays missing. Nothing is interpolated. Breaks in series, provisional values and low-reliability cells keep their Eurostat flags and are marked on the charts.
- Observations and estimates are kept apart. Fitted lines are labelled as fits; associations are not presented as causes.
- Every chart has a table. The “Show data table” button under each figure gives the same numbers as text, for screen readers and for anyone who wants to copy them.
- Synthetic means synthetic. The conjoint piece is flagged at the top, in the figures and in the data file.
How the data were retrieved and checked is described in the colophon.