There is an unfair burden on the self-employed. Walk the evidence below.
A main-occupation self-employed person owes 20.5% of net taxable income, with a legal minimum of about €890/quarter (whatever they earn) and a cap near €5,100/quarter. The minimum is owed even in a loss quarter, unless a formal exemption applies. "Over-contributing" means paying more than 20.5% of profit - which happens whenever the minimum exceeds 20.5% of a low profit, or provisional payments were set too high.
Sources: INASTI/NISSE annual report 2024 (1,299,825 self-employed; 801,544 main occupation); SPF Economie / INASTI (€5.234bn total contributions in 2023, +9.7% vs 2022); World Bank (self-employed ≈15% of employment); rates & minima per Belgian social insurance funds (Securex, Partena, Acerta), 2025-2026. Averages are means (pulled up by high earners); the typical contribution is lower.
Self-employed who made a loss or paid more than 20.5% of their revenue in social contributions are contributing more than expected.
Hair & beauty specialists are the most affected, with 74% of them paying more than the expected rate.
*unless a formal exemption applies (complementary activity, students, or a low-income waiver).
29% of the self-employed who earn less than €10k profit pay more than 50% of their profit in social contributions. Most self-employed sit in those profit bands - exactly where the share paying over half their profit in SC peaks.
Burden and revenue are decoupled: the self-employed with less than €10k profit per year only contribute 16% to the total social contributions, while being exposed to the largest individual impact (29% of them pay more than 50% of their profit in social contributions).
Each bar is a profit band's share of all social contributions paid; the line is the share of people in that band paying more than half their profit. Tap to toggle either series.
Of those over the expected rate, the split between overspenders and people pinned to the flat legal minimum, by band. The minimum dominates at the bottom.
The minimum is the single biggest driver of over-contribution at the bottom. This models the revenue effect of changing it. Only people the floor currently binds (loss-makers and the lowest earners) are affected - everyone already paying 20.5% of profit is untouched. Move the sliders.
Floor-only model: lowering the minimum reduces revenue; it does not yet model revenue-neutral progressivity (higher top-bracket rates). Sample of valid BE sole-trader years scaled x87.6 so the baseline reproduces the national €5.234bn (SPF Economie / INASTI, 2023); excludes company directors. Approximation - floored people are scaled to the new minimum rather than tracked individually.
A sole trader owes social contributions on their whole profit. A BV/SRL owner pays personal SC only on the director salary they choose to draw - the rest of the profit escapes contributions entirely. Set a profit and a salary below to see the gap for one company, then the revenue a reform charging SC on the full profit would raise across Belgium.
For one company, at this profit and salary
Across all ~300,000 management companies in Belgium
2025 rates, excl. fund fees. SC = 20.5% of the base (14.16% above ~€72.8k), legal minimum ~€3,562/yr, capped at ~€20,412/yr. Company owner pays SC on the director salary (capped at profit); retained profit faces corporate tax (~20-25%) instead, not shown - this illustrates the SC gap only, not total tax burden. National figure assumes ~300,000 single-director management companies (conservative; ~647,000 BV/SRL registered in total, CBE). Static model - ignores behavioural response and corporate-tax interaction, so it is an upper bound. Note the floor effect: at low profit the minimum already catches everyone, so the reform raises little; the escape hatch pays off at higher profits.
A fairer design would point the other way - easing the burden where it falls hardest so people can afford to start and grow, and applying the same logic across every status so the rules feel consistent whether you are a sole trader or a company director. The numbers above suggest these two moves could broadly pay for each other.