The evaluators counted participation cycle by cycle: 7,736 people in the pilot (2,138 at 27 assemblies, 251 delegates, 5,985 voters, four districts), 13,889 in Cycle 2 (eight districts), 18,184 in Cycle 3 (1,661 at 69 assemblies, 333 delegates, 16,642 voters, ten districts, then the largest PB vote in the US) and 51,362 voters with 606 active delegates and 6,127 assembly participants in Cycle 4 (24 districts). Later figures are the Council's own: over 67,000 voters in Cycle 5, over 102,000 in Cycle 6, 99,252 in Cycle 7 (27 districts), 118,308 in Cycle 8 (32 districts), and nearly 130,000 in the 2025-26 cycle (22 districts), which the Council called its highest turnout; no voter counts are given for 2023 or 2025. Who took part: women were over 60 percent at every stage; in the pilot, District 8 voters with incomes under $10,000 were 22 percent against 4 percent of 2009 election voters, and Latino voters 50 percent against 39 percent; in Cycle 3, 62 percent of voters were people of color, 39 percent had incomes under $35,000 (21 percent of 2013 election voters), 36 percent were foreign-born, 14 percent non-citizens, 6 percent aged 16 to 17, and 68 percent had never worked with others on a community problem; in Cycle 4, 57 percent were people of color and 23 percent could not vote in regular elections. What got funded: schools and libraries dominated (21 of 55 winners in Cycle 3; 80 percent of voters picked a school or library project), then parks; the 2018, 2019 and 2026 results lists are still mostly school technology, bathrooms, gyms, playgrounds, libraries, bus countdown clocks, trees and NYCHA upgrades; District 8 and 33 voters funded public housing projects; the pilot produced a library vending machine and a composting system. Delegates with incomes under $25,000 reported large gains in comfort contacting agencies (50 percent) and public speaking (38 percent); 82 percent said they were more likely to join a community organization. What the City changed: members funded $935,000 of non-winning or ineligible ideas in the pilot and over $2.8 million in Cycle 3, and Council Member Lander credited PB demand for the Department of Education doubling school bathroom funding from $50 million to $100 million. Negatives the evaluators found: participation dipped in most returning districts, most voters touch the process only at the ballot, delegate retention was under half, online channels skewed white and affluent, and participants reported frustration at implementation delays. Public Agenda counted NYC as 24 of the 46 PB processes in the US and Canada in 2014-15 and 28 of 61 in 2015-16. Celina Su's 2020 article on shifts in spending toward schools, streets and public housing is in the source list but was link only and not read.