Income and wealth inequality are related but distinct phenomena. The Nordic countries present a striking paradox: they have among the lowest income Gini coefficients in the developed world, but among the highest wealth concentration.
Several mechanisms explain this:
Funded pensions: Countries with large public pension systems (like Sweden’s AP funds) accumulate wealth collectively rather than individually. Private pension wealth is concentrated among higher earners.
Universal services: When healthcare, education, and childcare are publicly provided, lower-income households have less need for precautionary savings. This paradoxically increases measured wealth inequality.
Leveraged homeownership: High homeownership rates combined with mortgage debt mean many households have low net wealth despite owning property.
Tax treatment: Historically favorable tax treatment of capital income and wealth in some Nordic countries (Sweden abolished its wealth tax in 2007).
import sys
sys.path.insert(0, '..')
import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import matplotlib.ticker as mtick
import seaborn as sns
from data.config import FOCUS_COUNTRIES, EXTENDED_COUNTRIES, COUNTRY_NAMES
from data.plotting import (set_style, country_color, country_name,
save_figure, annotate_countries)
from data import wid, swiid
set_style()This scatter plot places income inequality on the x-axis and wealth inequality on the y-axis. If income and wealth inequality were perfectly correlated, all countries would lie on the diagonal. Instead, the Nordics cluster in the lower-left to upper-left area: low income Gini, high wealth concentration.
scatter_countries = ['US', 'GB', 'FR', 'DE', 'SE', 'DK', 'NO', 'FI',
'NL', 'IT', 'ES', 'CA', 'AU', 'AT', 'CH']
# Income Gini from SWIID
income_gini = swiid.get_gini(scatter_countries, gini_type='disp',
start_year=2017, end_year=2022)
# Wealth: top 10% share from WID as a proxy for wealth concentration
wealth_top10 = wid.get_wealth_shares(
scatter_countries,
percentiles=['p90p100'],
start_year=2017, end_year=2022,
)
if not income_gini.empty and not wealth_top10.empty:
# Take most recent per country
ig = (income_gini.sort_values('year', ascending=False)
.groupby('country_code').first().reset_index())
wt = (wealth_top10.sort_values('year', ascending=False)
.groupby('country').first().reset_index())
wt = wt.rename(columns={'country': 'country_code', 'value': 'wealth_top10'})
merged = ig.merge(wt[['country_code', 'wealth_top10']], on='country_code')
if not merged.empty:
fig, ax = plt.subplots(figsize=(10, 8))
colors = [country_color(c) for c in merged['country_code']]
ax.scatter(merged['gini'], merged['wealth_top10'],
c=colors, s=120, zorder=5, edgecolors='white', linewidth=1)
annotate_countries(ax, merged['gini'].values,
merged['wealth_top10'].values,
merged['country_code'].values)
ax.set_xlabel('Disposable income Gini (SWIID)')
ax.set_ylabel('Top 10% wealth share (WID)')
ax.yaxis.set_major_formatter(mtick.PercentFormatter(1.0))
ax.set_title('Income Inequality vs. Wealth Concentration (~2020)\n'
'Nordic paradox: low income Gini, high wealth concentration',
fontsize=13)
plt.tight_layout()
save_figure(fig, '04_income_vs_wealth_gini')
plt.show()
else:
print('Could not merge income and wealth data.')
else:
print('Income Gini or wealth share data not available.')
Wealth concentration has its own dynamics. In some countries, the top wealth shares declined during the 20th century (war, taxation, inflation eroding capital) and have been rising since the 1980s. In others, the pattern differs.
ts_countries = ['US', 'FR', 'GB', 'SE', 'DE', 'NO']
for pct, pct_label in [('p90p100', 'Top 10%'), ('p99p100', 'Top 1%')]:
wealth_ts = wid.get_wealth_shares(
ts_countries, percentiles=[pct], start_year=1970, end_year=2023
)
if not wealth_ts.empty:
fig, ax = plt.subplots(figsize=(12, 6))
for cc in ts_countries:
data = wealth_ts[wealth_ts['country'] == cc].sort_values('year')
if not data.empty:
ax.plot(data['year'], data['value'],
color=country_color(cc), linewidth=2,
label=country_name(cc))
ax.yaxis.set_major_formatter(mtick.PercentFormatter(1.0))
ax.set_xlabel('Year')
ax.set_ylabel(f'{pct_label} wealth share')
ax.set_title(f'{pct_label} Wealth Share (1970-present)', fontsize=14)
ax.legend()
plt.tight_layout()
save_figure(fig, f'04_wealth_{pct}_timeseries')
plt.show()
else:
print(f'Wealth share data not available for {pct_label}.')
Are income and wealth concentration moving together? In some countries, rising top income shares have translated into rising wealth concentration. In others, the relationship is weaker.
for cc in ['US', 'SE', 'FR']:
income_ts = wid.get_top_shares_timeseries(
[cc], percentile='p90p100', concept='pretax', start_year=1970
)
wealth_ts = wid.get_wealth_shares(
[cc], percentiles=['p90p100'], start_year=1970, end_year=2023
)
if not income_ts.empty and not wealth_ts.empty:
fig, ax1 = plt.subplots(figsize=(12, 5))
ax2 = ax1.twinx()
ax2.spines['right'].set_visible(True)
ax2.spines['right'].set_color('#cccccc')
inc = income_ts[income_ts['country'] == cc].sort_values('year')
wth = wealth_ts[wealth_ts['country'] == cc].sort_values('year')
ax1.plot(inc['year'], inc['value'], color='#4878a8', linewidth=2,
label='Top 10% income share')
ax2.plot(wth['year'], wth['value'], color='#c75b5b', linewidth=2,
linestyle='--', label='Top 10% wealth share')
ax1.yaxis.set_major_formatter(mtick.PercentFormatter(1.0))
ax2.yaxis.set_major_formatter(mtick.PercentFormatter(1.0))
ax1.set_xlabel('Year')
ax1.set_ylabel('Top 10% income share', color='#4878a8')
ax2.set_ylabel('Top 10% wealth share', color='#c75b5b')
ax1.set_title(f'{country_name(cc)}: Income vs. Wealth Concentration',
fontsize=14)
lines1, labels1 = ax1.get_legend_handles_labels()
lines2, labels2 = ax2.get_legend_handles_labels()
ax1.legend(lines1 + lines2, labels1 + labels2, loc='upper left')
plt.tight_layout()
save_figure(fig, f'04_income_vs_wealth_ts_{cc}')
plt.show()
else:
print(f'Data not available for {country_name(cc)}.')
The bottom 50% typically owns very little wealth — often close to zero or even negative (when debts exceed assets). This is true even in relatively egalitarian countries.
bottom50_wealth = wid.get_wealth_shares(
ts_countries, percentiles=['p0p50'], start_year=1970, end_year=2023
)
if not bottom50_wealth.empty:
fig, ax = plt.subplots(figsize=(12, 6))
for cc in ts_countries:
data = bottom50_wealth[bottom50_wealth['country'] == cc].sort_values('year')
if not data.empty:
ax.plot(data['year'], data['value'],
color=country_color(cc), linewidth=2,
label=country_name(cc))
ax.axhline(y=0, color='gray', linestyle='--', alpha=0.5)
ax.yaxis.set_major_formatter(mtick.PercentFormatter(1.0))
ax.set_xlabel('Year')
ax.set_ylabel('Bottom 50% wealth share')
ax.set_title('Bottom 50% Wealth Share (1970-present)\n'
'The bottom half owns very little everywhere', fontsize=14)
ax.legend()
plt.tight_layout()
save_figure(fig, '04_wealth_bottom50_timeseries')
plt.show()
else:
print('Bottom 50% wealth share data not available.')
A visual decomposition of why wealth inequality can be high even when income inequality is low.
fig, ax = plt.subplots(figsize=(12, 7))
mechanisms = [
'Funded pension\nsystems',
'Universal public\nservices',
'Mortgage-heavy\nhomeownership',
'Capital income\ntax treatment',
'Compressed wages\n(low income Gini)',
]
effects_on_income = [0, 0, 0, 0.1, -0.5]
effects_on_wealth = [0.3, 0.4, 0.3, 0.2, -0.1]
x = np.arange(len(mechanisms))
width = 0.35
bars1 = ax.barh(x + width/2, effects_on_income, width,
label='Effect on income inequality',
color='#4878a8', alpha=0.8)
bars2 = ax.barh(x - width/2, effects_on_wealth, width,
label='Effect on wealth inequality',
color='#c75b5b', alpha=0.8)
ax.set_yticks(x)
ax.set_yticklabels(mechanisms)
ax.axvline(x=0, color='black', linewidth=0.5)
ax.set_xlabel('Direction of effect on measured inequality\n'
'(negative = reduces inequality, positive = increases it)')
ax.set_title('The Nordic Wealth Paradox: Why Low Income Inequality\n'
'Coexists with High Wealth Concentration', fontsize=14)
ax.legend(loc='lower right')
ax.annotate('These mechanisms reduce the\nneed for private savings,\n'
'concentrating measured wealth\namong those who do save.',
xy=(0.35, 3.5), fontsize=9, fontstyle='italic',
bbox=dict(boxstyle='round,pad=0.5', facecolor='#fef9e7', alpha=0.8))
plt.tight_layout()
save_figure(fig, '04_nordic_paradox_mechanisms')
plt.show()
Income and wealth inequality are distinct. Countries with low income inequality can have high wealth concentration, and vice versa.
The Nordic paradox is real. Sweden, Denmark, and Norway have among the highest wealth concentration in the OECD despite low income inequality.
Public provision reduces private accumulation. Universal services and public pensions reduce the need for precautionary savings, paradoxically increasing measured wealth inequality.
Wealth concentration has been rising. In most countries, top wealth shares have increased since the 1980s, though patterns vary by country.
The bottom 50% owns very little everywhere. This is a near-universal feature of wealth distributions, even in egalitarian societies.