Cyber Income Inequality

We study the income inequality among streamers using the administrative data of a leading Chinese live-streaming platform. The live-streaming technology enables a superstar to produce new entertainment products matched with demand and occupies a larger market share. Imagine an extreme case; the best streamer hosts live for 24 hours, earns all possible income, and leaves zero time for other streamers. Our data show that the income distribution of the highest-paid streamers follows Zipf’s Law and appears to be even more concentrated than any offline business: NBA top players, Forbes celebrities, and billionaires. Income inequality increased rapidly as the platform expanded from 2018 to 2020 — for example, the income share of the platform’s top 10 streamers increased from 14.82% to 45.15% as its revenue grew by 142%. To estimate inequality elasticity to the market size, we study four quasi-experimental shocks: potential market size proxied by economic development and Fintech coverage, quarter-end revenue spikes induced by the seasonal incentive regime, user surge induced by capital raising, and the Covid-19 lockdown in Wuhan. Gini coefficient elasticity ranges from 1.3% to 10.6%, estimated from the cross-city variations (local economic development and Covid-19 Wuhan lockdown); the time-series variations (quarter-end and user surge before capital raising) imply an elasticity ranging from 3.6% to 25.5%