AcknowledgmentsHundreds of people helped with this project and I am deeply indebted to all of them. First, this project only exists because of the incredible women and men who work on the streets and put in extra hours to organize their colleagues and demand a better life. I am grateful to the dozens of street vendors in Bolivia and Brazil who decided to trust me, talk with me, and teach me. I am especially grateful to the women and men who took the time to explain how the markets work, how to sell in them, and how to interact with officials as a vendor. I regret that I can't thank people by name for confidentiality reasons, but from the bottom of my heart, thank you G.for teaching me how to set up a stall and be a vendor, and I hope we have many more conversations about politics over api. Thank you to the champañeras for making me an honorary sister and for letting me spend so many happy hours with all of you in the plaza. Thank you R. for being my sócio, teaching me how to set up and protect a stall, and sharing all of your stories with me; let's plan for many more. Thank you to everyone at the Federation and the Markets Office for your patience, your stories, and your advice. In addition to the vendors that are part of this project, the project would not have gone very far without the help of Cecilia Salazar, Luciana Molina

Bolivia is one of the poorest countries in Latin America with a gross domestic product of around US$3500 per capita, health spending of approximately $220 per capita, a labour market dominated by informal work, and a weak health system. However, in the response to COVID-19, Bolivia has fared better than other health systems in the region and provides insight with regard to the implementation of subnational nonpharmaceutical interventions and supporting workers without social protection.The Bolivian Government confirmed the first case of COVID-19 in the country on March 10, 2020, and responded quickly by cancelling events, closing schools and borders, and implementing a national lockdown on March 22, 2020. However, the Bolivian Government was under pressure to open the economy in an election season. In response, the Bolivian Government shifted responsibility for most non-pharmaceutical interventions to departmental and municipal governments on June 1, 2020. The Bolivian Government maintained a mask mandate, school and border closures, and a nightly curfew, while allowing departmental and municipal governments to set workplace, social gathering, population mobility, and public transit policies. Daily deaths from COVID-19 increased markedly from 20 on June 1, 2020, to 96 on Aug 1, 2020. 1 Subnationally, the first outbreak of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) occurred in the wealthy, politically conservative department of Santa Cruz, followed by the poor, politically moderate department of Beni in May, 2020. These two departments maintained strict confinement policies in June, 2020, to manage outbreaks. Strict policies were also maintained in the wealthy, politically conservative department of Tarija, while the other six departments relaxed restrictions quickly in June, 2020, and subsequently a surge in cases was observed. In response to an increase in deaths in June and July, 2020, Chuquisaca and Cochabamba reinstated some restrictions and the other departments shifted to a policy of isolating municipalities with outbreaks. Daily deaths began to decline in August, 2020, decreasing to 35 per day by Sept 15, 2020. 1 We declare no competing interests.

Reformers claim that public subsidies and regulation of political finance reduce corruption in politics, while observers worry that they have no impact on corruption, or even increase it. Despite national-level debates and billions of dollars spent, few studies have tested this relationship. The authors argue that political finance reform mitigates corruption by reducing private money's importance in politics and increasing the sanctions for corrupt behavior. Elite interviews from Paraguay's political finance reform illustrate the argument and elaborate the theoretical mechanisms. The study evaluates the argument using an original dataset measuring political subsidies from 175 countries from 1900–2015, as well as disaggregated corruption measures from the Varieties of Democracy project. The findings support the thesis that political finance reform reduces corruption, even in countries where such reforms are unevenly implemented.

IntroductionTo present an analysis of the Brazilian health system and subnational (state) variation in response to the COVID-19 pandemic, based on 10 non-pharmaceutical interventions (NPIs).Materials and methodsWe collected daily information on implementation of 10 NPI designed to inform the public of health risks and promote distancing and mask use at the national level for eight countries across the Americas. We then analyse the adoption of the 10 policies across Brazil’s 27 states over time, individually and using a composite index. We draw on this index to assess the timeliness and rigour of NPI implementation across the country, from the date of the first case, 26 February 2020. We also compile Google data on population mobility by state to describe changes in mobility throughout the COVID-19 pandemic.ResultsBrazil’s national NPI response was the least stringent among countries analysed. In the absence of a unified federal response to the pandemic, Brazilian state policy implementation was neither homogenous nor synchronised. The median NPI was no stay-at-home order, a recommendation to wear masks in public space but not a requirement, a full school closure and partial restrictions on businesses, public transportation, intrastate travel, interstate travel and international travel. These restrictions were implemented 45 days after the first case in each state, on average. Rondônia implemented the earliest and most rigorous policies, with school closures, business closures, information campaigns and restrictions on movement 24 days after the first case; Mato Grosso do Sul had the fewest, least stringent restrictions on movement, business operations and no mask recommendation.ConclusionsThe study identifies wide variation in national-level NPI responses to the COVID-19 pandemic. Our focus on Brazil identifies subsequent variability in how and when states implemented NPI to contain COVID-19. States’ NPIs and their scores on the composite policy index both align with the governors’ political affiliations: opposition governors implemented earlier, more stringent sanitary measures than those supporting the Bolsonaro administration. A strong, unified national response to a pandemic is essential for keeping the population safe and disease-free, both at the outset of an outbreak and as communities begin to reopen. This national response should be aligned with state and municipal implementation of NPI, which we show is not the case in Brazil.