Cities- Their economic contribution in relation to their population share
- Neural City Team
- Sep 26
- 11 min read
Neural City Data Blog
Cities are where economies come alive. They concentrate talent, infrastructure, and opportunities—and in doing so, shape national growth trajectories.
But how concentrated should growth be? Are some countries more dependent on a handful of metros than others?
To explore this, we compared the top 3, 5, and 10 highest-GDP cities in the world’s four largest economies: the United States, China, Germany, and India. We then looked at how much of the national GDP these cities produce relative to the population they host.
Each country uses slightly different “city” definitions.
U.S. numbers are for entire metropolitan areas
China’s include wide administrative territories (so Chongqing looks very large)
Germany uses city proper (so populations/GDP look smaller)
and India uses metros/urban agglomerations.
The figures are not strictly apples-to-apples, but useful for showing concentration patterns.
What We Did
City Selection: For each country, we identified the largest metros by GDP (e.g., New York, Los Angeles, and Chicago in the US; Shanghai, Beijing, and Shenzhen in China; Berlin, Hamburg, and Munich in Germany; and Delhi, Mumbai, and Bengaluru in India).
Grouping: We calculated GDP and population shares for the top-3, top-5, and top-10 cities in each country.
Comparison: By dividing GDP share by population share, we assessed how much more (or less) economic output these cities generate compared to their population weight.
Early Insights
India’s metros are extreme outliers. The top-10 cities account for only ~9% of the population but nearly 28% of GDP, a ratio of ~3×—far higher than the US (~1.5×), Germany (~1.5×), or China (~1.8×).
Urbanisation differs sharply. Despite India’s visibly crowded metros, a far smaller share of Indians live in big cities compared to the US or China. This means India’s growth is being carried by relatively fewer urban residents.
Social vs spatial inequality diverges. India has the lowest Gini coefficient among the four countries, suggesting relatively equitable incomes. Yet, spatially, its growth is the most concentrated—signaling that inequality looks very different when measured across people vs across places.
The following three graphs together reveal not just how much cities drive national output, but also how India’s story diverges sharply from global peers.
This bar graph compares how much more economic output the top 3, 5, and 10 cities in each country generate relative to the share of people they host.
If the ratio = 1 → cities contribute GDP equal to their population share.
If >1 → cities punch above their weight in economic terms.
Country snapshots
Germany & USA: Ratios hover around 1.4–1.6. Their biggest cities produce a bit more GDP than their population share, but the difference is modest.
China: Starts high (2.25 for top-3) but flattens quickly (~1.78 by top-10). Economic growth is distributed more evenly across provinces and mid-size cities.
India: Stands out. Ratios are ~3 for top-3, top-5, and top-10. This means India’s top metros contribute nearly 3× more GDP than their share of population—a level of concentration unmatched by peers.
Why India’s Difference Is Huge
At first glance, the gap between 3 (India) vs 1.5–1.8 (others) seems like “just 1–1.5 points.” But in absolute dollar terms, it’s enormous:
India’s nominal GDP FY 2022–23 ≈ ₹272 lakh crore (~$3.3 trillion).
Top-10 cities: 27.6% of GDP ≈ ₹75 lakh crore (~$900 billion).
Their population share: 9.3% of Indians (~133 million people).
If India’s metro concentration looked more like China’s (~1.8× ratio), their GDP contribution would be closer to ₹45–46 lakh crore.
The extra ~₹29–30 lakh crore (~$350–360 billion) concentrated in metros is equal to the entire economy of a mid-sized G20 country.
The Indian bars on the chart matter: the “small” numeric difference hides a huge real impact.
Indian Perspective
Growth Concentration Risk: With ~3× over-representation, metros like Delhi, Mumbai, Bengaluru, Chennai, and Hyderabad are carrying a disproportionate share of the economy. Any infrastructure breakdown here (floods, pollution, transport collapse) hits national growth.
About the graph
X-axis: Share of national population living in the top 3, 5, or 10 cities.
Y-axis: Share of national GDP produced in those same cities.
Each country has 3 dots (Top-3, Top-5, Top-10), joined by a dotted line.
The 45° diagonal would represent “GDP share = Population share”. Points above it mean cities are producing more GDP relative to their population.
What can we learn from this graph
1. India’s Low Urbanisation
India’s dots sit far left compared to the USA, China, and even Germany.
That means India’s top cities host a smaller share of the national population.
Despite cities being visibly crowded, the country as a whole is less urbanised.
By contrast, the USA has much higher urbanisation: its top cities alone account for 15–25% of the population.
2. India’s Steep Slope
India’s line rises much more sharply than other countries.
This means for every extra % of population that lives in big cities, India gets disproportionately more GDP.
Put differently: Indian metros are economic overachievers, producing far more output per person than the rest of the country.
3. Comparing Peers
Germany: Sits closer to the 45° line → its cities produce GDP roughly in line with their population share. Economic activity is more evenly spread.
China: Higher than Germany, but its slope flattens as you go from Top-3 to Top-10, showing that secondary cities are also pulling weight.
USA: High population shares, high GDP shares, but slope is more balanced. The growth is distributed across many urban regions.
India: Small population share, big GDP share → extreme concentration.
Why It Matters for India
National growth depends on metros. Delhi, Mumbai, Bengaluru, Chennai, and Hyderabad drive a vastly outsized share of GDP.
Infrastructure risk: Because economic weight is so concentrated, any disruption in these metros—transport bottlenecks, climate shocks, flooding—has a magnified national impact.
Policy blind spot: Gini coefficients and welfare transfers may make India look “equitable” on income distribution, but spatial inequality in growth is stark.
Gini coefficient is calculated from household income (or consumption) surveys, not from where GDP is produced.
In India, most households are still in rural or semi-urban areas, and their consumption levels are relatively similar at a low base. This flattens the distribution curve and produces a lower Gini.
In contrast, the GDP concentration analysis we did is about geography of production (metros vs rest). These are two different slices:
Functional inequality: Between households across the whole population → Gini.
Spatial inequality: Between cities vs. the rest → our GDP share vs population ratio.
India’s Gini coefficient is the lowest (0.255) → income inequality looks better than China, USA, or Germany.
This mismatch highlights that national inequality metrics miss the spatial story. India’s welfare schemes and subsidies compress consumption differences at household level, but metros still carry outsized economic weight.
Role of socialist & welfare state policies
India has a long tradition of redistributive policies: food subsidies (PDS), employment guarantees (MGNREGA), free school meals, farm loan waivers, DBT transfers, etc.
These programs compress consumption inequality (people across rural/urban areas have more similar minimum consumption baskets), which pushes the Gini downward.
In OECD terms, these are like in-kind transfers and subsidies that reduce observed gaps, even though underlying productivity is highly unequal.
Germany & USA
Both countries have lower city concentration ratios (1.4–1.6) → their growth is spread more evenly across regions.
Yet, the US has the highest Gini (0.375) → inequality shows up at the household level, not in spatial concentration.
Germany has both low spatial concentration and low income inequality, a sign of more balanced development.
China
City concentration is moderate (1.7–2.2), and Gini (0.357) is higher than India’s but lower than the US.
China’s pattern shows: strong urban centres but also large provincial economies, so the gap between cities and the rest isn’t as extreme as India’s.
Why This Matters for India
India’s metros = economic engines, but they are also single points of failure.
National inequality stats (Gini) can give a misleading sense of “equity” while masking the fact that growth is geographically narrow.
Data: Sources, Notes, Caveats
Here’s the list of the top 3, 5, and 10 GDP-leading cities/metros we used in the analysis for each of the four countries.
India (FY 2022–23, nominal GDP, metros/urban agglomerations)
Top 3: Delhi, Mumbai, Bengaluru
Top 5: +Chennai, Hyderabad
Top 10: +Kolkata, Ahmedabad, Pune, Surat, Jaipur
China (2023, nominal GDP, city administrative units)
Top 3: Shanghai, Beijing, Shenzhen
Top 5: +Chongqing, Guangzhou
Top 10: +Chengdu, Hangzhou, Wuhan, Nanjing, Suzhou
Germany (2021, nominal GDP, city proper)
Top 3: Berlin, Hamburg, Munich
Top 5: +Frankfurt, Cologne
Top 10: +Hanover Region, Stuttgart, Düsseldorf, Nuremberg, Bremen
United States (2023, nominal GDP, metropolitan statistical areas / MSAs)
Top 3: New York, Los Angeles, Chicago
Top 5: +San Francisco Bay Area, Dallas–Fort Worth
Top 10: +Washington DC, Houston, Boston, Atlanta, Seattle
Note on comparability:
Each country uses slightly different “city” definitions.
U.S. numbers are for entire metropolitan areas
China’s include wide administrative territories (so Chongqing looks very large)
Germany uses city proper (so populations/GDP look smaller)
and India uses metros/urban agglomerations.
The figures are not strictly apples-to-apples, but useful for showing concentration patterns.
City GDP Figures
Country | GDP Top 3 Cities % | Pop. Top 3 Cities % | (GDP% / Pop. %) Top 3 Cities | GDP Top 5 Cities % | Pop. Top 5 Cities % | (GDP% / Pop. %) Top 5 Cities | GDP Top 10 Cities % | Pop. Top 10 Cities % | (GDP% / Pop%) Top 10 Cities |
China | 10.3 | 4.57 | 2.25 | 15.1 | 8.18 | 1.85 | 23.3 | 13.12 | 1.78 |
Germany | 11.9 | 8.44 | 1.41 | 15.8 | 10.69 | 1.48 | 22.2 | 14.17 | 1.57 |
India | 14.1 | 4.72 | 2.99 | 19.4 | 6.29 | 3.08 | 27.6 | 9.28 | 2.97 |
United States | 20.1 | 12.5 | 1.61 | 26.2 | 16.3 | 1.61 | 37.9 | 25 | 1.52 |
United States (Metropolitan areas, nominal USD, 2023)
China (City administrative units, nominal USD, 2023; NBS-based)
Shanghai — $729.5B
Beijing — $672.0B
Shenzhen — $495.3B
Guangzhou — $436.0B
Chongqing — $434.4B Wikipedia
Germany (City proper, nominal EUR, 2021)
Berlin — €165.46B
Hamburg — €130.87B
Munich — €128.75B
Frankfurt am Main — €74.09B
Cologne (Köln) — €66.69B(Most recent consolidated city-level table is 2021.) Wikipedia
India (Metropolitan areas, nominal USD, FY 2022–23)
Delhi NCR — $195.6B
Mumbai (MMR) — $170.69B
Bengaluru — $118.13B
Chennai — $96.18B
Hyderabad — $84.13B Wikipedia
Rank | Metro (MSA) | USD (bn) |
1 | New York–Newark–Jersey City | 2,298.9 |
2 | Los Angeles–Long Beach–Anaheim | 1,295.4 |
3 | Chicago–Naperville–Elgin | 894.9 |
4 | San Francisco–Oakland–Berkeley | 778.9 |
5 | Dallas–Fort Worth–Arlington | 744.7 |
Source: BEA metro GDP 2023 table (values listed are 2023 dollars as reported by BEA for real GDP).
Rank | City | RMB (bn, 2023) | USD (bn, 2023 FX) |
1 | Shanghai | 5,140.4 | 721.8 |
2 | Beijing | 4,735.4 | 664.9 |
3 | Shenzhen | 3,490.3 | 490.1 |
4 | Chongqing | 3,061.4 | 429.9 |
5 | Guangzhou | 3,072.4 | 431.4 |
Sources: 2023 city GDP list; 2023 average FX from NBS (7.1217 RMB per USD).
China (converted from 2023 RMB using CNY/USD avg 2023 = 7.1217)
Rank | City | EUR (bn, 2021) | USD (bn, 2023 FX) |
1 | Berlin | 165.46 | 179.8 |
2 | Hamburg | 130.87 | 142.3 |
3 | Munich | 128.75 | 139.9 |
4 | Frankfurt am Main | 74.09 | 80.5 |
5 | Cologne (Köln) | 66.69 | 72.5 |
Sources: 2021 city GDP table; 2023 EUR per USD ~0.92 (World Bank WDI table 4.16). Wikipedia+1
Germany (converted from 2021 EUR using 2023 avg EUR/USD ≈ 1.087)
Rank | Metro Area | INR (bn, 2022–23) | USD (bn, 2023 FX) |
1 | Delhi NCR | 15,375.37 | 186.2 |
2 | Mumbai (MMR) | 13,417.15 | 162.5 |
3 | Bengaluru | 9,285.52 | 112.5 |
4 | Chennai | 7,560.56 | 91.6 |
5 | Hyderabad | 6,613.10 | 80.1 |
Sources: 2022–23 metro GDP list; 2023 INR/USD average from FRED (AEXINUS).
India (converted from FY 2022–23 INR using 2023 avg INR/USD = 82.5708)
Nominal GDP and Base Year
China
Base year: None in the same sense as BEA.
China’s National Bureau of Statistics (NBS) reports GDP of cities in current prices (nominal, RMB) each year.
There are also “constant price” (real) GDP growth rates, but those are not published at the city level the way US chained dollars are.
So the Shanghai, Beijing, Shenzhen etc. figures we used are nominal 2023 RMB, not inflation-adjusted.
Germany
Base year (for real GDP): Destatis uses 2015 = 100 as the reference year in their national accounts.
But for city-level GDP, the most accessible tables (like Berlin, Hamburg, Munich, etc.) are usually published in current prices (nominal EUR).
So the 2021 Berlin/Hamburg numbers we converted are nominal EUR 2021, not chained. If you needed real growth, they’d link it back to 2015 prices.
India
Base year: India’s national accounts use 2011–12 as the base year for real GDP at constant prices (both national and state).
But city/metropolitan GDP is not officially published by CSO/MoSPI. Instead, think tank/economic surveys compile them—almost always in current prices (nominal INR).
The Delhi, Mumbai, Bengaluru, etc. metro figures we used are nominal INR 2022–23.
Caveats
Geography differs: U.S./India use metropolitan areas; China figures are for city administrative units (which can include vast rural tracts); Germany list is city proper. Apples-to-apples is hard—use these as directional comparisons. Wikipedia+3
Years differ slightly: U.S. (2023), China (2023), Germany (2021), India (FY 2022–23).
Source notes: U.S. values align with BEA metro GDP; China values reflect NBS-reported city GDP compiled in the list; India values are a compiled metro list with cited government/economic survey sources per city row
Germany recency: 2021 is the latest consistent city-level compilation we could source; newer 2023 data exist at NUTS-2/metro-region level, but not as a uniform city list. The USD conversions here use 2023 FX to match your “USD 2023” requirement. Wikipedia+1
India definitions: The Indian list is metropolitan areas (UA/FUA)—not just municipal limits—and is for FY 2022–23. Conversions use calendar-year 2023 FX for consistency. Wikipedia+1
China FX: We used the NBS-stated 2023 average of 7.1217 RMB per USD for conversion; that keeps it official and consistent across cities. National Bureau of Statistics of China
Method Notes
US: BEA metro GDP for 2023 (these are real/chained dollars per BEA’s table; still in “USD 2023,” just inflation-adjusted) Wikipedia
China: 2023 city GDP in RMB → USD with 2023 average CNY/USD = 7.1217 (NBS communiqué) Wikipedia+1
Germany: city GDP for 2021 in EUR (latest city-level official compilation) → USD with 2023 average EUR/USD (World Bank WDI 2023 shows ~0.92 EUR per USD, i.e., 1 EUR ≈ 1.087 USD) Wikipedia+1
India: metro GDP FY 2022–23 in INR → USD with 2023 average INR/USD = 82.5708 (FRED annual) Wikipedia+1
Gini coefficient
(0 = perfect equality, 1 = perfect inequality)
Country | Gini (value) | Year | Definition / Source |
United States | 0.375 | 2022 | Disposable income, after taxes & transfers (OECD Income Distribution Database). (OECD) |
China | 0.357 | 2021 | World Bank PIP “Gini index” (household survey–based; WB series). (World Bank Open Data) |
Germany | 0.295 | 2024 | Equivalised disposable income, after taxes & transfers (Eurostat; as reported by TE pulling Eurostat). (Trading Economics) |
India | 0.255 | 2022 | World Bank PIP “Gini index” (WB update; see also GoI press note citing the WB figure). (World Bank Open Data) |
Population
United States - 2023
City / Metro (MSA) | 2023 Population (m) | Share of U.S. pop |
New York–Newark–Jersey City | 19.498 | ~5.8% |
Los Angeles–Long Beach–Anaheim | 12.886 | ~3.9% |
Chicago–Naperville–Elgin | 9.263 | ~2.8% |
San Francisco–Oakland–Berkeley | 4.567 | ~1.4% |
Dallas–Fort Worth–Arlington | 8.100 | ~2.4% |
Sources. Metro populations from ACS 2023 (NYC, Chicago, SF) and FRED/Census series for LA; DFW from Census-based 2023 estimate noted by Wikipedia. National denominator = U.S. resident population 2023 (Vintage 2024 estimates). Census.gov+5Census Reporter+5Census Reporter+5
China - 2023
City | 2023 Resident Pop (m) | Share of China pop |
Shanghai | ~24.87 | ~1.8% |
Beijing | 21.858 | ~1.6% |
Shenzhen | ~17.66 | ~1.3% |
Guangzhou | 18.827 | ~1.3% |
Chongqing | ~32.05 | ~2.3% |
Sources. National denominator: 1,409.67m at end-2023 (NBS). City populations from municipal communiqués/official summaries: Beijing 21.858m; Guangzhou 18.827m; widely cited 2023 figures for Shanghai (~24.9m), Shenzhen (~17.7m), and Chongqing (~32.1m). MacroTrends+3National Bureau of Statistics of China+3Beijing G
Germany - 2021
City | 2021 Population (m) | Share of Germany pop |
Berlin | ~3.68 | ~4.4% |
Hamburg | ~1.85 | ~2.2% |
Munich | ~1.49 | ~1.8% |
Frankfurt am Main | ~0.79 | ~0.9% |
Cologne (Köln) | ~1.08 | ~1.3% |
India - 2021
Metro / UA | ~2023 Pop (m) | Share of India pop |
Delhi UA | 32.94 | ~2.3% |
Mumbai UA | 21.30 | ~1.5% |
Bengaluru | 13.61 | ~0.9% |
Chennai (Metro/UA) | 11.78 | ~0.8% |
Hyderabad | 10.80 | ~0.8% |
Caveats
Matching GDP geography:
U.S. uses Metropolitan Statistical Areas (MSAs);
China uses city administrative units (include large rural counties for Chongqing);
Germany is city proper;
India uses metros/urban agglomerations (UA).Keep that in the caption so readers don’t try to compare “city proper” to “metro.” National Bureau of Statistics of China
Denominators:
U.S. national population: Vintage 2024 Census estimates table;
China national: NBS end-2023;
Germany national (for 2021): Destatis;
India national (for 2023 to align with FY22–23 city GDP year): UN/WPP.
Comments