The population pyramid is the canonical demographic supply chart — negative values for one gender enables direct visual comparison of the gender balance within each age cohort. Aging of peak cohorts (35–44) signals medium-term attrition risk. Gender gaps within age bands reveal structural participation barriers.

Dual line (M vs F on same axis) makes the gender gap immediately visible as the vertical distance between series. The 30-year view captures structural shifts (EU accession, crisis, recovery) that shorter windows miss. Area fill for female series emphasises the historically suppressed participation level and its recent upward trend.

Stacked bar with composition adds up to 100% at each period, revealing structural upgrade of the employed workforce. The shrinking low-education share and growing high-education share is the education premium in action. Cross-reference with wages to quantify the premium per ISCED level.

Long-term unemployment (12+ months) signals skills erosion and social exclusion — a qualitatively different problem from frictional unemployment. Colour-coding the bars by severity turns a neutral distribution into an urgency map. The 37% in 2+ years is the policy intervention target.

Sectoral composition is the structural baseline for any workforce planning — it tells you the size of each labour pool. Donut works here because there are only 4 major categories. Cross-reference with OJA sector demand data to detect mismatches: a large supply sector with low OJA activity may signal displacement risk.

Inactivity reasons reveal non-participation potential — the "hidden workforce" that could re-enter with the right policy. Grouped bars by gender on the same chart expose the care penalty: women's inactivity is dominated by care duties (38%) vs men's by disability and retirement. These are different policy problems requiring different interventions.

The Beveridge curve is the canonical macro-labor chart — it distinguishes cyclical from structural unemployment. Movement along the curve is cyclical (recovery/recession). An outward shift of the curve signals structural mismatch: more vacancies coexisting with more unemployment than before. Greece's curve has shifted outward post-2020 — a matching efficiency problem.

V/U ratio by occupation separates genuine skill shortages (high V, low U) from general unemployment concentrations. A ratio >1 means more vacancies than unemployed in that field — employers compete for candidates. <0.5 means many more unemployed than openings — a supply-side problem. The vertical reference line at 1.0 makes the comparison instant.

Dual Y-axis is necessary here because OJA postings and workforce count differ by orders of magnitude. The diverging trend — postings growing 4× while workforce grew <15% — is the core finding. A single-axis chart would obscure either the scale of demand growth or the stability of supply. The bar+line type visually differentiates the two series without requiring a legend.

Diverging bar at a zero baseline is the canonical skills gap chart — it shows both direction (shortage vs surplus) and magnitude in one view without mental arithmetic. Traditional side-by-side bars require the reader to compute the difference. Ordering by gap severity puts the highest-priority training investments at the top and bottom.

Points above the parity diagonal have OJA-posted salaries below actual market wages — employers are under-posting, risking rejection. Points below indicate over-posting — inflated offers that may not materialise. This systematic comparison is only possible by merging two normally separate datasets and reveals systematic OJA salary disclosure bias by occupation.

Sorting by normalised rate (per 1k LF) removes country-size bias and reveals the true intensity of OJA activity relative to market size. Colour coding by intensity tier provides immediate visual grouping. Tooltip shows both the rate and raw counts — always necessary when presenting normalised metrics to avoid misleading impressions about market scale.

Geographic scatter using real lon/lat coordinates gives spatial intuition without requiring GeoJSON registration. Bubble size encodes OJA intensity; colour encodes employment rate — two variables readable simultaneously. This "pseudo-map" approach is fully portable (no map tiles, no CORS) and works in any browser. For a full choropleth, register Eurostat NUTS GeoJSON with echarts.registerMap().

Indexed series removes absolute scale differences and focuses attention on relative acceleration. COVID shock visible in Q2 2020; Data Scientist demand shows steepest climb driven by GenAI adoption post-2022.

Grouped horizontal bar enables direct side-by-side reading per skill. Horizontal layout handles long skill labels. Ordered by demand descending — most critical gaps at top.

Bubble chart encodes 4 variables simultaneously. Roles top-right with large bubbles are high-demand, high-paying, fast-growing — the hardest to hire. Instantly prioritises sourcing budget.

Calendar heatmap (skills × month) reveals seasonal patterns invisible in yearly averages. Useful for planning recruitment campaigns: knowing Cloud roles spike in Q1 means sourcing campaigns should launch in November–December.

Funnel chart quantifies where candidates drop at each stage — the stage with the steepest drop is the process bottleneck. Unlike a bar chart, the funnel shape makes conversion implicit and the comparison to the stage above immediate. Essential for hiring process optimization.

Lollipop (stem + dot) is a cleaner alternative to a bar chart for median comparison — less ink, more focus on the value. Sorting by descending duration makes the hardest-to-fill roles immediately visible. The dot's precision contrasts with bar's mass, appropriate for a single-point estimate like the median.

Donut shows channel mix composition. Cross-referencing with cost-per-hire data (tooltip) enables ROI analysis — referrals typically have low cost and high retention despite small volume. The key insight is usually that the highest-volume channel is not the highest-ROI channel.

Diverging bar at zero baseline shows both direction (surplus/shortage) and magnitude simultaneously, without the mental arithmetic required by side-by-side bars. The sector filter makes this parametrized: planners can focus on their sector without losing the full-spectrum reference. At 18 occupations the chart height scales to remain readable.

Radar is uniquely suited for comparing two multivariate profiles on identical axes. The area between polygons is the skills gap — larger area means more intensive intervention needed. Surplus areas (supply exceeds demand) suggest over-training opportunity costs.

Sankey is the canonical flow chart — link width encodes volume, revealing structural misalignments (e.g. engineering graduates not entering tech) and unexpected talent pools. No other chart type shows multi-to-multi flows readably.

Waterfall decomposes net change into contributing factors, each shown running from the prior cumulative value. It answers "why did employment in Software Development grow by X?" — impossible to do with any other chart type. Identifies which lever (graduation rates, retention, immigration) has most impact.

Internal age pyramid mirrors the LFS supply pyramid but applied to a specific organisation's workforce. Aging of peak cohorts signals medium-term attrition risk. Gender gaps within bands reveal internal equity issues distinct from market supply gaps.

Grouped bar (current vs target) is the most direct way to show attainment against a plan. The gap between bars is the open roles count — readable without annotation. This chart drives staffing committee conversations and priority-setting for TA resources.

Stacked area shows both absolute total trend AND departmental composition shift simultaneously. Attrition spikes (drop in a layer) and growth spurts (expanding layer) are immediately visible. Stream-style (smooth area) is appropriate when evolution is gradual; use stepped bar for discrete hiring events.

Force-directed network reveals both skill importance (node size) and cluster topology simultaneously. Bridge nodes (between clusters) are the highest-value, most versatile skills. No bar chart or heatmap can show graph topology. Drag interaction allows exploratory navigation.

Bump chart plots rank (not absolute value) on the Y-axis, making crossing lines the story. Superior to line charts for ranking data because absolute values vary in scale across skills. Answers "which skill overtook which?" — the key longitudinal research question.

Treemap uses area to show part-of-whole relationships across a hierarchy. Where a pie handles ~6 slices before becoming unreadable, treemaps handle 30–50 items. It respects the hierarchical structure of ESCO naturally — colour encodes parent cluster, size encodes demand volume.

Scatter with regression tests the scarcity pricing hypothesis: growing, scarce skills command wage premiums. Outliers above the line are "underpriced" — potentially a training investment opportunity before the market corrects. Bubble size adds supply volume without cluttering.

Sunburst is the interactive drill-down alternative to a treemap — radial layout groups hierarchy levels visually by ring depth. Clicking a sector zooms into that branch. Ideal for researchers navigating a taxonomy: start at the ESCO group level, drill into the specific skill cluster of interest.

Parallel coordinates compare multivariate profiles across occupations simultaneously. Lines that stay close to each other indicate similar skill profiles — occupation families. Crossings reveal where occupations differ sharply on a specific skill. No radar can compare more than 2–3 occupations readably.

S-curves reveal the diffusion stage of each skill: skills in exponential phase are the highest training priority (early mover advantage). Skills in plateau (Python) still matter but training ROI is declining. Skills just emerging (LLMs/GenAI) are in the steepest exponential phase — typical S-curve inflection point.

Small multiples (same chart type, identical axes, repeated for each country) allow cross-country comparison without the overplotting of multi-line charts with 6+ series. Shape differences (slope, volatility, recovery speed) are read by eye-scanning. Standard format in Eurostat and OECD publications.

Bubble matrix for strategic portfolio analysis: risk (x), scale of impact (bubble = employment), economic weight (colour = wage) are the three policy dimensions. Roles top-right with large bubbles are high-automation-risk, high-employment — the priority target for reskilling policy. Roles bottom-left are policy tailwinds.

Stacked area shows composition AND total simultaneously. A rising share of fixed-term contracts alongside rising total employment is a precarity warning signal. The zero baseline makes absolute changes visible alongside proportional shifts.

Annotated timeline overlays policy event markers on a KPI trend line, making causal reasoning explicit. Without annotations, a reader can only observe "the rate fell" — with annotations they see "the rate fell 3pp following Youth Guarantee expansion." Standard format in OECD effectiveness reports and Cedefop monitoring publications.

Slope chart (line between exactly 2 time points) is cleaner than a bar chart for showing change when there are multiple entities. The slope's angle encodes magnitude and direction simultaneously. Unlike a full time-series line chart, it focuses attention on "before vs after" rather than intermediate fluctuations — ideal for policy evaluation comparisons.

Grouped bar (male/female per country) makes the within-country gap visible as the space between adjacent bars. Countries are ordered by female employment rate to rank policy performance. This is preferable to a dumbbell chart when absolute levels (not just the gap) carry policy meaning — Denmark's near-equal bars are structurally different from Italy's large gap even if the gap size were the same.

NUTS2 is the EU policy-planning unit — the level at which Structural Funds are allocated. Sorting by employment rate with a national average reference line identifies below-average regions (policy priority) and above-average regions (can absorb workers from elsewhere). This chart feeds directly into ESIF programming decisions.

Box plot is the most information-dense single chart type for a distribution: median, IQR, tails, and outliers in one compact symbol. A single "average salary" figure is misleading — knowing the full distribution (P25–P75 IQR) is actionable career information. Multiple grouped boxes enable cross-occupation comparison.

Sankey encodes transition volume as link width — the most natural mapping for flow data. Replaces anecdote ("I heard developers move into data science") with empirical evidence ("45% of developers who change occupation enter data science"). Wide links indicate high-probability, well-worn transition paths.

Line + confidence interval band makes uncertainty explicit. For career decisions, knowing demand will "probably be 20–35% higher in 2028" is more actionable than a spuriously precise "27% higher." Widening band over time correctly represents increasing uncertainty. Standard format in Cedefop Skills Forecast publications.

Personal skills radar gives the job seeker an immediate gestalt of readiness profile. The area enclosed by "your skills" vs "market demand" is the total gap. Unlike a table of scores, the radar shows which gaps are large (require significant upskilling) vs small (easily bridged). Widely used in career development tools (LinkedIn, Coursera) — users already know how to read it.

Grouped bar by seniority level across occupations shows both the career earnings premium for progression AND cross-occupation comparison at each level. The slope of bars within an occupation is the seniority premium — steep slopes favour staying deep in a specialisation rather than switching; flat slopes suggest switching for higher pay is rational.

Gauge chart gives an immediate single-number gestalt of readiness — the most common UX pattern in career tools, fitness apps, and risk dashboards. It works precisely because it simplifies: users act on a score of 68 differently than they'd act on a table of 8 skill sub-scores. The colour band (red → amber → green) adds urgency encoding without additional explanation.

Grouped horizontal bar (current vs required per skill) is the operational training roadmap. Horizontal layout handles long skill labels. Gap between bars directly encodes training intensity needed per skill. Unlike a diverging bar, this layout keeps both absolute levels visible — knowing that ML is at 42/100 current vs 88/100 required tells you both the gap AND the starting competence level.