The effects split by era. In the Workers' Party years the sources agree on redistribution (Marquetti found that the poorer a region the more investment per head it received in 1989 to 2000, a result its authors qualify because it uses planned rather than executed budgets; Baiocchi found no link between a region's investment and its Workers' Party vote or associational density), on delivery (execution averaged 96 percent up to 2000; 95.9 percent of the 5,556 demands of 1990 to 2004 were eventually built), on services (households with water from 49 to 98 percent and with sewage from 46 to 85 percent between 1989 and 1996, 900 km of sewers against 1,100 km in all previous administrations, 30 km of paving a year, school enrolment doubled, 28,862 families given housing assistance in 1992 to 1995 against 1,714 in 1986 to 1988, attribution to the OP being the authors' judgment rather than a measured counterfactual), on growth (under 1,000 participants in 1990, about 17,000 in 2002, over 20,000 in 2015, more than 18,000 in 2025, and about a fifth of residents had attended at least once by 2006) and on who came. The city's associations grew from about 300 to 540 between 1988 and 1998, fastest in the poorest districts, and half the activists Baiocchi surveyed had started in the OP; the 2009 survey, by contrast, found a majority of participants with no associational tie at all. The learning effects are what founders, participants and the Núñez study testify to; the World Bank found budget literacy still low and Fedozzi's studies found a critical citizenship consciousness in only a minority even after eight years. The property tax went from 5.8 percent of revenue in 1990 to 13.8 percent in 1992 after OP leaders occupied the chamber to back the reform. The negative column belongs to 2002 onwards: a backlog that reached 2,395 unexecuted demands by 2017 (the oldest from 1994), with only 8.6 percent of the 2013 to 2016 demands built; satisfaction with the City's accounting falling from 80.7 percent in 1995 to 39.4 percent in 2015; council renewal collapsing after unlimited re-election in 2008; and a secretary of works removed from office in 2012 for using OP works as an electoral machine. The World Bank's cross-city econometrics found poverty lower and water and sanitation access higher in Brazilian cities that had run PB since before 1996, but no consistent gain in fiscal performance and a higher share of current spending; Spada's later panel found the effects concentrated in the first four years and larger on the mayor's party's re-election than on spending. The world-model effect is double-edged: the UN listed it among 40 best practices in 1996, over 80 Brazilian cities had copied it by 1998 and thousands worldwide since, and that prestige is the main reason evaluators give for why no mayor has dared abolish it.