The pooled model places 77 officials who cast votes on 2,081 divided actions from 2000 through May 2026, including those decided over a single dissent. Use the council selector to explore individual members, or hover and zoom on the comparisons. Party labels were researched independently of the votes and never enter the model. Party labels set only the starting values of the two estimation chains. Two frequent lone dissenters, Jolanda Jones and Helena Brown, anchor the ends of the scale. See the technical appendix for estimation and identification details.
With at least seven qualifying votes
5 mayors
From elections in 1999–2023
Both panels use the same axes and smoothing bandwidth. Curves stop at each group’s most extreme observed members; mixed-affiliation Vasquez is omitted. These small samples are descriptive, not population distributions.
Select a council to update the member positions in place. Hover for a name, affiliation, estimate and conditional interval; click a point to keep its details below. Use the legend to show or hide a party and drag across the score axis to zoom.
The new package adds substantial early coverage. All member positions come from the pooled fit, so recurring officials keep the same coordinate. Intervals describe uncertainty within the reported solution.
The five at-large members and the mayor were elected citywide; eleven members represent districts. At-large members sit closer to Whitmire as a group, but at-large member Plummer is the council’s second most negative member, after Pollard. This is a descriptive 2023-elected-council comparison, not proof that citywide elections cause moderation. Official February 2024 seat roster.
The purple diamond marks the median of the five at-large members’ pooled positions and uses Houston’s approximate citywide Trump share, calculated by summing the allocated votes across all eleven districts. It is not a twelfth district or an input to the district-only trend line. Split precincts make the citywide vote estimate approximate too.
Party affiliation and presidential results are external checks, not inputs to the fit. Houston municipal ballots are nonpartisan. Presidential votes are allocated to council districts by precinct-area share, so the district comparison is approximate. Mayors and at-large members have no single district.
This dashboard uses published Houston roll-call votes from 2000 through May 2026 and separately researched party-affiliation evidence. The model is fitted to 2,081 divided actions and gives each included official a fixed position. Both chains start from researched party labels and agree closely; the fit is corroborated by a separately fitted 2021–2026 window. An earlier diagnostic found that a dispersed random start can reach a different orientation of the historical end, as the technical appendix records.
The expanded report explains the findings and limitations. The technical appendix gives data construction, estimation, diagnostics, and evidence sources. A Texas common-space bridge diagnostic explains why the six available legislative bridges do not yet support a credible city-to-state map. An earlier 2021–2025 report is retained for comparison.