---

title: Brazilian coauthor candidates status: shortlist (2026-06-01, updated 2026-06-26)

Brazilian coauthor candidates

The plan calls for a Brazilian coauthor with:

• Prior work on pesquisas eleitorais specifically (not generic voting behavior). Published or WP; Brazilian-journal articles count equally with international journals. The criterion is "has thought hard about this specific data and its institutional context", not "has cleared a top-5 desk."
• Early-career stage — PhD student, postdoc, or junior faculty.
• Poli sci or econ publication orientation.
• Brazilian residence or strong Brazilian network so they can validate the institutional setting AND promote / present the paper inside Brazil.

2026-06-26 update — ranking changes

Web verification of current affiliations changed two of the original five rankings:

• Guilherme Russo (was rank 1) → dropped to "courtesy contact only". Russo has finished the Vanderbilt PhD and now serves as Director of Research at Quaest Pesquisa e Consultoria while continuing as a tutor-professor at FGV-EESP and a researcher at FGV-CEPESP. Quaest is one of the polling firms whose behavior the paper studies — structural conflict of interest, same category as Felipe Nunes. Moved to the "Disqualified" section below.
• Ryan Lloyd (was rank 4) → moved out of academia. The USP postdoc was a prior role; he is now a Policy Specialist (Trust & Safety, LatAm/Brazil) / UX Researcher at Google, after also spending time as Visiting Assistant Professor at Centre College. He no longer satisfies the "early-career academic" criterion and the publication incentive on his side is weaker. Kept on the list at rank 4 as a domain-expert collaborator (two of the most relevant Brazil polling papers in print — Lloyd-Turgeon 2021 IJPOR + lloyd2016vote with Gramacho) but no longer the obvious rank-1.

Working ranking after the update:

1. Lucas Gelape (UFMG) — only candidate satisfying all four criteria cleanly (Brazil-resident, assistant prof on tenure track, top-Brazil polisci department, methodology fit). No polling publication yet — this paper would be his entry.
2. Frederico Batista Pereira (UNC Charlotte) — closest topical fit (competing-mechanism paper on the same data); mid-career, US-based.
3. Pedro Santos Mundim (UFG) — SECOM-insider angle; mid-career; runs a postgrad program in Brazil.
4. Ryan Lloyd (Google) — strongest topical fit on paper, but in industry now; coauthor only if he wants to keep a publication track alive on the side.

Detailed entries below in the new ranking order.

1. Lucas Gelape — Assistant Professor, UFMG (Political Science); PhD USP 2021

Why: Brazilian, in Brazil, at UFMG (the strongest Brazilian polisci department for electoral methodology — Felipe Nunes's home before FGV). Recent PhD (USP 2021), postdoc at FGV-CEPESP 2023–2025, visiting fellow at Harvard. Specializes in municipal elections, electoral geography, candidate behavior — exactly the units of analysis this paper uses. Also has applied data-analysis experience with Brazilian news organizations (Volt Data Lab, Núcleo Jornalismo, G1), which means he's institutionally connected to the polling-and-journalism ecosystem that consumes our results. Affiliated with CEL-UFMG, FGV-CEPESP, NEPOL-UFJF, REPSAL, and MAPE (IESP-UERJ).

Fit on the four criteria:

• Prior pesquisa-eleitoral publication: △ his electoral work is heavy on vote distributions / electoral geography / campaign financing, not specifically polls (verified 2026-06-26 from lgelape.github.io/en/research/). If we coauthor, this paper would be his entry to the polling sub-literature.
• Early-career: ✓ assistant professor, PhD 2021.
• Poli sci / econ orientation: ✓ at a top Brazilian PhD department.
• Brazilian promotion: ✓✓ best of the four for promotion inside Brazil — he's there, at UFMG.

Risks / things to verify: none specific to flag in the 2026-06-26 re-check — his profile is consistent with the 2026-06-01 notes.

Likely path in: direct email; cite his municipal-elections / electoral-geography work in the intro and emphasize the spatial / neighborhood-coverage angle of the design (coverage_class extraction from DS_DADO_MUNICIPIO).

2. Frederico Batista Pereira — Assistant Professor, UNC Charlotte (Political Science); PhD Vanderbilt 2016; UFMG BA/MA

Why: Coauthor of Batista Pereira & Nunes (2024) Batista Pereira & Nunes 2024 in Opinião Pública — the competing-mechanism paper on the same TSE-registered data (argues poll-vs-result gap is late voter movement, not poll error). Brazilian by training (UFMG BA/MA), now US-based since Vanderbilt PhD. Research focuses on voter behavior, public opinion, political information and misinformation — adjacent enough that polling methodology is natural for him.

Fit:

• Prior pesquisa-eleitoral publication: ✓ the most-relevant alternative paper to ours.
• Early-career: △ mid-career — PhD 2016, ~10 years out, Assistant Professor.
• Poli sci / econ orientation: ✓ poli sci, but publishes in JPolitics and Latin American Public Opinion Review.
• Brazilian promotion: △ US-based at UNC Charlotte; strong Brazilian network (Nunes / UFMG ties) but not resident in Brazil.

Risks / things to verify: his coauthorship history with Nunes (Quaest insider) — if Nunes is a frequent collaborator, the team overlap is structurally informative but possibly delicate given the "sponsor-bias" framing of our paper. Worth surfacing in the first conversation.

Likely path in: approach with explicit acknowledgment of the sibling-paper relationship — "your 2024 paper argues mechanism X; we want to test mechanism Y on the same data; we don't think the papers compete, they complement; want to collaborate?"

3. Pedro Santos Mundim — Associate Professor, UFG (Federal University of Goiás); Coordinator of the Postgraduate Program in Political Science

Why: Authored Mundim et al. (2019) Mundim, Vidigal & Michelotti 2019 in Revista do Serviço Público — a survey-experimental study on opinion polls about the Bolsa Família program, commissioned by the Brazilian government's SECOM. Mundim served as Special Advisor and Director of Public Opinion Research Assessment at SECOM (Communication Secretariat of the Presidency) 2014–2015, where he coordinated 57 quantitative + 11 qualitative surveys. So he has first-hand experience as both the sponsor and the analyst of government- commissioned polls — institutional knowledge of how sponsor-commissioned polls actually operate in Brazil that no other candidate on this list has.

Fit:

• Prior pesquisa-eleitoral publication: ✓ Brazilian-journal paper on public-opinion polls; SECOM working experience.
• Early-career: △ Associate Professor (PhD IUPERJ/IESP-UERJ 2010, ~15 years out), so mid-career; coordinates a postgrad program, so has institutional bandwidth.
• Poli sci / econ orientation: ✓ poli sci.
• Brazilian promotion: ✓ in Brazil, runs the UFG postgrad program; active in Opinião Pública / Brazilian journal ecosystem.

Special value: the insider-sponsor perspective is the strongest defense against accusations of naïveté in the "polls are biased" framing. Mundim has been on the sponsor side and can speak to which Channel A choices actually get made in practice — a unique methodological asset for the paper.

Likely path in: approach with the Channel A / Channel B framing — "you've been the sponsor; we want to measure the effect of being a sponsor. Help us not get the institutional context wrong."

4. Ryan Lloyd — Policy Specialist / UX Researcher, Google; PhD Texas-Austin 2016

Why: Coauthor of Lloyd & Turgeon 2021, "Polling in New Democracies and Electoral Malpractice: The Case of Brazil" in International Journal of Public Opinion Research 33(4):1039– — the single most thematically relevant prior paper ("electoral malpractice" in Brazilian polls is the closest existing framing to sponsor bias). Also coauthor on Lloyd et al. (2016) (Lloyd, Turgeon & Gramacho on vote buying and polling error). Two polls papers, both on Brazil.

Current status (2026-06-26). Lloyd has left academia: he is now a Policy Specialist (Trust & Safety, LatAm/Brazil) and User Experience Researcher with the Engineering Productivity Research team at Google, after also spending time as Visiting Assistant Professor at Centre College and a postdoc at the USP IRI. He is no longer on the academic tenure track — so the publication-incentive side of a coauthor pitch is weaker than for the other three candidates, but the topical fit and prior Brazil-polling track record are the best on the list.

Fit on the four criteria:

• Prior pesquisa-eleitoral publication: ✓✓ two highly relevant papers.
• Early-career: ✗ no longer in academia.
• Poli sci / econ orientation: ✓ historical record is poli sci / IJPOR.
• Brazilian promotion: △ at Google (LatAm Trust & Safety implies some Brazil presence) but not in a Brazilian university; cannot do the domestic seminar / journal-network promotion that the criterion prioritizes.

Risks / things to verify: bandwidth and motivation — academic publishing as an industry researcher is usually a side project, so turnaround is slower and his deliverable share has to be light. May also be useful as a consultant / acknowledgments-only contributor on the Brazilian-polling-history sections rather than a full coauthor.

Likely path in: approach with the framing "we don't need much from you — your IJPOR paper is the closest analog, and we'd like your read on the institutional history sections + sponsor-bias positioning. Open to coauthorship if it fits, or to an acknowledgment + early-reader role if not." Lower the ask up front so the industry-bandwidth concern is already handled.

Disqualified as coauthor (industry conflict)

Guilherme Russo — Director of Research, Quaest; FGV-EESP/CEPESP affiliation

Was rank 1 in the 2026-06-01 notes on the strength of Meireles & Russo 2022 (Meireles & Russo (2022)). 2026-06-26 web verification finds he has finished the Vanderbilt PhD and now serves as Director of Research at Quaest Pesquisa e Consultoria while continuing as a tutor-professor at FGV-EESP and a researcher at FGV-CEPESP. Quaest is one of the polling firms in our analysis sample — same structural conflict of interest as Felipe Nunes (Quaest CEO/founder). Treat as Quaest-channel courtesy contact, alongside Nunes, for the pre-publication heads-up. Not a coauthor candidate.

Nuance — Quaest looks clean in our results. Worth flagging that the COI here is optics, not material incentive distortion. By AN-085 (docs/analyses/an-085-trusted-source-decomposition.md), Quaest is one of five "trusted firms" (with DATAFOLHA, PARANÁ PESQUISAS, REAL TIME MÍDIA, VERITA) that the universe-wide spec identifies as cleaner than the rest: trusted-firm β = −1.15 pp on mean |error| (p=0.0004) and −2.02 pp on margin error (p<0.002); trusted-firm self-contracted polls are the most accurate slice of the entire dataset (5.2 pp mean |error| vs universe 8.1 pp). AN-087 notes Datafolha + Quaest are driving the trusted-firm coefficient. So Russo's interest isn't "skew results in Quaest's favor" — by the current estimates, Quaest is already favored on the empirics. The remaining case for disqualification is the appearance one: a Quaest insider on the byline of a paper that vindicates Quaest is exactly the result-laundering pattern poli-sci editors flag. If Henrik decides the disclosure footnote is enough, Russo's intellectual contribution would be unusually high (the Meireles & Russo 2022 paper is our most direct predecessor). The decision is judgment, not a hard rule.

Other names checked 2026-06-26 (not coauthor fits)

• André Borges — Associate Professor, UnB (Institute of Political Sciences). Parties / party systems / electoral behavior, with a subnational focus on Latin America. Coeditor of The Recasting of the Latin American Right (CUP 2024) with Ryan Lloyd and Gabriel Vommaro — so Lloyd's UnB-side network anchor. Too senior + not poll-specific for a coauthor role, but a natural referral source for UnB PhD students working on polling methodology, and a useful warm-intro channel if a Lloyd outreach (as industry collaborator) needs an academic-side bridge.
• Natália Bueno — Associate Professor, Emory (Political Science). Strong Brazil-elections + misinformation record with top-5 publications (APSR, AJPS, BJPS, JoP). Research is field-experimental on welfare provision, race in elections, and misinformation effects; not pesquisa-eleitoral methodology. Useful citation; not a coauthor fit.

Senior gateways (referral sources, not target coauthors)

• Wladimir Gramacho — Associate Professor, UnB (Faculdade de Comunicação). PhD Salamanca. Author of Gramacho (2013) and Gramacho (2015), the original Brazilian poll-accuracy audits. Runs the Centro de Pesquisa em Comunicação Política e Saúde Pública (CPS / UnB). Too senior for the user's criteria as a coauthor, but his PhD students at UnB are the natural pool to ask him about — he's been training Brazilian poll-research talent for over a decade. A brief outreach asking "is anyone among your current orientandos working on polling methodology and interested in a coauthorship?" is high-yield. Also a connector to Lloyd via the lloyd2016vote coauthorship — useful if Lloyd's industry move means his orientandos / former students at Texas-Austin or USP are a better channel than Lloyd himself.
• Rachel Meneguello — Full Professor, Unicamp; Director of CESOP (Centro de Estudos de Opinião Pública), which runs the Estudo Eleitoral Brasileiro (ESEB) post-election panel. The CESOP doctoral pipeline is the second natural pool for orientando referrals; same ask as Gramacho.
• Felipe Nunes — Associate Professor, UFMG + faculty at FGV-EESP; CEO and founding partner of Quaest. Conflict-of-interest as a coauthor, but the most academically active member of the Brazilian polling industry. Worth a courtesy heads-up before the paper goes out so it isn't read as an attack on Quaest, and as a path to referrals among his graduate students at UFMG/FGV who might have polling interests independent of his industry role.

Brazilian research hubs to query for PhD students

The user's relaxed criterion (WPs and Brazilian-journal papers count) opens up the doctoral pipeline at three centers:

• CECOMP-UFMG (Centro de Estudos do Comportamento Político) — large group of profs, postdocs, and PhD students from UFMG (Felipe Nunes's old home; Frederico Batista Pereira's MA/BA institution). Highest-density quantitative-electoral group in Brazil; the post-Nunes generation is precisely the demographic the user wants. Gelape (cand 1) is now there and can refer.
• CESOP-Unicamp — runs ESEB, the Brazilian electoral survey. PhD students working on survey methodology there are direct fits.
• CEPESP-FGV (São Paulo) — Russo (now Quaest) and Gelape (cand 1, postdoc 2023–25) are both alumni; current PhD students publish WPs through CEPESP-Data and have access to the CepespData platform that already integrates TSE data.

Notes for the outreach pass

• Only Lloyd has published pre-election polling papers on Brazil, but he's now in industry. The four-criteria sweet spot is no longer occupied — Gelape needs a polling entry; Lloyd needs an academic motive; Batista Pereira and Mundim are mid-career not early. The pragmatic plan is Gelape as primary coauthor + Lloyd as a light-touch collaborator on the polling-history sections.
• The paper's framing must avoid a "Brazilian polls are crooked" narrative — anyone with Nunes / Quaest / FGV-CEPESP ties (cand 2 via Nunes; Russo via direct Quaest employment) will be sensitive to that, and even cand 1 (Gelape) is in the news-org data ecosystem. The Channel A / Channel B decomposition in summary.md § Mechanism is the politic framing: "we document average sponsor effects; whether they're persuasion or fabrication is what the decomposition reveals."
• The 2026-06-01 advisory "don't approach more than one of Russo and Batista Pereira simultaneously" is now moot — Russo is off the list. Gelape-first, with Batista Pereira and Lloyd as parallel light-touch follow-ups, is safe.

TODOs from this section

• First contact: Lucas Gelape at UFMG. Direct email pitching him as the polling-literature-entry collaborator; emphasize Channel A / coverage-class spatial angle for fit with his electoral-geography track record.
• Light-touch parallel contact: Ryan Lloyd at Google. Position as polling-history reviewer / acknowledgment role, with coauthorship open if he wants it. Read Lloyd & Turgeon 2021 IJPOR first (queued in todo.md for campus download).
• Read Meireles & Russo (2022) (PDF in _pdfs/) — for the methodological positioning of the paper, regardless of Russo's coauthor status.
• Read Batista Pereira & Nunes (2024) (PDF in _pdfs/) — Batista Pereira's mechanism argument; figure out where our paper stands relative to theirs (still load-bearing whether or not he becomes a coauthor).
• Decide whether to send Gramacho a brief "do you have an orientando interested" email before approaching Gelape — useful backup channel given the polling-publication gap on Gelape's CV.
• Same Gramacho-style ask to Meneguello at CESOP.
• Check Mundim et al. (2019) — Mundim's 2019 Bolsa-Família paper — to assess whether the SECOM-insider perspective is usable as a paper section (institutional context box) regardless of whether he ends up a coauthor.
• If a CECOMP-UFMG PhD student turns out to be the right fit, Gelape is the natural intra-Brazil bridge — he'd know who's currently working on polls inside the group.