Wealth Town (3D)

400 households trade at random: towers grow with wealth. Test a tax and a basic income.

Switch to 2D

Presets

Playback

Market

Policy

Camera

Layers

Quality

Model check

The science

What you see

Every plot is one household. The tower on it grows with its wealth (floors = 0.8 + 2.2 w0.8, roof colour on hover), the colour of the floors is the same wealth scale as the 2D page, flying coins are single trades, and a burst of big coins to the treasury at the top of the map is a tax round (shown for the eight richest households). The Blender-modelled town only draws the numbers; the economy runs in the same code as the 2D page.

What is modelled

400 households start with the same average wealth (the unit of the page: 1 = the average household). Each trade picks two random households. Between trades nothing is created or destroyed, so the total stays exactly 400; only the way it is shared changes. After every N trades (one cycle) a tax round runs.

Equations

  • Equal-split market (Dragulescu and Yakovenko 2000): the pair pools s = wi + wj and re-splits it at random, wi' = u s, wj' = (1-u) s, with u uniform in [0,1]. The stationary distribution is the exponential law P(w) = e-w (mean 1), whose Gini index is exactly 1/2 and whose Lorenz curve is L(p) = p + (1-p) ln(1-p).
  • Yard-sale market (Chakraborti 2002, Boghosian 2014): the pair bets a share f of the poorer side, d = f min(wi, wj), on a coin toss. The poor-favouring bias is P(i wins) = 1/2 + (beta/2)(wj - wi)/(wi + wj). With a fair coin (beta = 0) every household's expected wealth is constant, yet the distribution condenses: a few households end up with nearly everything.
  • Tax and basic income: each cycle a share tau of the wealth above theta x average is collected and paid out equally to all households. It conserves the total exactly, and with enough tax the Gini index settles at a finite value instead of drifting to 1.
  • Inequality measures: G = 2 Σ i w(i) / (n Σ w) - (n+1)/n for wealths sorted upward; the Palma ratio is the top-10 % share divided by the bottom-40 % share; "poor" means below half of the median.

Checks run in the page

  • Exact conservation of total wealth through trades and taxes.
  • The equal-split market reaches Gini 0.5 and 63.2 % of households below the mean, as the exponential law predicts.
  • Gini of 1,1,3,3 equals 0.25 by hand, and two independent Gini formulas agree on the live state.
  • A taxed economy ends fairer than the same untaxed one.

Honest simplifications

  • Wealth only; no production, no wages, no prices, no inheritance and no debt. Real wealth distributions have a power-law tail for the richest and are not described by one rule alone.
  • "Tax" means the share is moved by a rule, with no behavioural response such as people avoiding the tax or working less.
  • One cycle is not a calendar period; the page is a model of how rules shape a distribution, not a forecast for any country.

What to try

  • Run "Pure yard-sale market" for a few hundred cycles and watch the top 10 % share climb although every toss is fair.
  • Raise the tax to 3 %, then 15 %, and read off where the Gini index settles.
  • Set the threshold to 3x so only the top is taxed ("Safety net") and compare the bottom-50 % share.
Loading model...
Loading Blender model...
Lorenz curve against the exponential law
Wealth histogram against exp(-w)
Gini and top / bottom shares over time
Made with 🟠 Blender · Three.js