Skip to main content

The new preserving and continuing Keith Poole’s infrastructure for scholars, students and observers of Congress


For the last 40 years, Keith Poole has developed and curated a trove of basic data and measurements related to the United States Congress. He has made these resources freely available through his widely-used website since 1995. At Poole’s, scholars, students, journalists, and the broader public could download gold-standard historical and current roll-call voting data, member rosters, NOMINATE scores, and measures and visualizations of party cohesion and polarization, among many other useful things. In this article, we describe how we are preserving and continuing these vast public-goods contributions through the new Developed and housed at UCLA, the new carries on the creation of basic roll-call data infrastructure, including the assignment of Poole’s widely-used ICPSR number-like identifiers to new members, data on every roll-call vote ever taken, NOMINATE scores and other standard roll-call vote-based measures such as party-loyalty scores. In addition to serving as a platform for the continuation and dissemination of this basic data infrastructure, the new also provides powerful tools for exploring the history of roll-call voting, the US Congress, and American politics and political history through a simple search interface and interactive visualizations.

This is a preview of subscription content, access via your institution.

Fig. 1
Fig. 2
Fig. 3
Fig. 4


  1. For examples of Poole’s substantive contributions see McCarty et al. (2016), Poole and Rosenthal (2000a), Poole (2007), McCarty et al. (2009) and McCarty and Poole (1995). For leading examples of Poole’s methodological contributions see Poole (2005), Poole (1998) and Poole and Rosenthal (1985).

  2. Poole’s also served as an informal repository for a variety of roll call voting data sets from legislative bodies other than the United States. Those data are included on our new site as part of our general archive of all Poole’s Our new site, as described here, is narrowly focused on voting in the United States Congress.

  3. A contested roll call is one for which neither the yeas nor the nays received over 97.5% of the vote.

  4. The url, accessed on January 11, 2018, yielded 1,091 results. A similar query on Google Scholar®–, accessed on January 11, 2018, yielded “about” 4040 results.

  5. Several additional NOMINATE algorithms have followed the original in the literature, including W-NOMINATE, D-NOMINATE, Alpha-NOMINATE, DW-NOMINATE and Nokken-Poole scores. provides DW-NOMINATE and Nokken-Poole scores. Users who want W-NOMINATE or Alpha-NOMINATE can easily estimate those on their personal computers using R packages that implement those procedures and the Rvoteview package described below to access roll-call data in R.

  6. The basis of the roll calls included in the database for the 1st though the 105th Congress is ICSPR Study #4; see

  7. ICPSR numbers are the member identifiers first assigned by the ICPSR and then extended by Poole to the current Congress.

  8. Modern Senate ICPSR numbers begin with ‘4’; House numbers begin with ‘2’, overflowing to ‘3’ as needed. The second and third digits indicate the year since 2000 beginning each Congress. For members who switch parties, in all appearances after the switch, the first digit is replaced replaced with ’9’.

  9. Because we do not allow the members’ locations to vary over time, our scores are based on a statistical model that is identical to that developed for W-NOMINATE. However, the way in which the data are organized and the model is fit is that used in DW-NOMINATE. Because we are providing what Poole called Common Space DW-NOMINATE scores, we call our scores “DW-NOMINATE” despite the fact the each members’ location is fixed over time.

  10. To be precise, the DW-NOMINATE algorithm will assign a unique ideal point to every unique ICPSR code, and members switching major parties or becoming president are the only times this happens. Examples of when a member is assigned a new ICPSR code without becoming president include Arlen Specter switching to the Democratic Party in 2009 and Strom Thurmond switching to the Republican Party in 1965.

  11. As with Poole’s original Fortran algorithm, the new algorithm is freely available on the web. It can be found at

  12. Furthermore, the hyper-parameters w and \(\beta\) described above are not updated regularly in this process. They are unlikely to change in meaningful ways in the short run, and to ensure long-run comparability of estimates, they will be re-estimated periodically—most likely at the end of each Congress.

  13. See

  14. Poole also regularly produced DW-NOMINATE estimates for the House and Senate separately that allowed the ideal points of long-serving members to evolve linearly over time. currently provides only Nokken-Poole scores for scholars interested in studying the changes in positions of members over time, and in the future we may provide other scores based on models that permit members’ ideal points to change over time.

  15. See

  16. See

  17. See

  18. The entire project is offered under the MIT public license, allowing for derivative works and extensions by interested users and scholars.


  • Carrol, R., Lo, J., Poole, K. T., & Rosenthal, H. (2009). Measuring bias and uncertatiny in dw-nominate using the parameteric bootstrap. Political Analysis, 17(3), 261–275.

    Article  Google Scholar 

  • Clausen, A. (1973). How congressmen decide: A policy focus. New York: St. Martin’s.

    Google Scholar 

  • Congress, U. S. (2016a). Legislative subject terms. Accessed January 22, 2017.

  • Congress, U. S. (2016b). Policy areas—Field values. Accessed January 22, 2017.

  • Lewis, J. B., DeVine, B., & Pritcher, Lincoln with Martis, K. C. (2013). United states congressional district shapefiles. Data files

  • Lewis, J., Boche, A., Rudkin, A. & Sonnet, L. (2017). rvoteview: Voteview Data in R. R package version 0.1—For new features, see the ’Changelog’ file (in the package source).

  • Martis, K. C. (1989). The historical atlas of political parties in the United States Congress, 1789–1989. Basingstoke: Macmillan Publishing Company.

    Google Scholar 

  • McCarty, N. M., & Poole, K. T. (1995). Veto power and legislation: An empirical analysis of executive and legislative bargaining from 1961 to 1986. Journal of Law, Economics, & Organization, 11, 282–312.

    Google Scholar 

  • McCarty, N. M., Poole, K. T., & Rosenthal, H. L. (1997). Income redistribution and the realignment of American politics. AEI studies on understanding economic inequality. Washington: AEI Press.

    Google Scholar 

  • McCarty, N., Poole, K. T., & Rosenthal, H. L. (2009). Does gerrymandering cause polarization? American Journal of Political Science, 53(3), 666–680.

    Article  Google Scholar 

  • McCarty, N., Poole, K. T., & Rosenthal, H. L. (2016). Polarized America: The dance of ideology and unequal riches. Cambridge: MIT Press.

    Google Scholar 

  • Nokken, T. P. & Poole, K. T. (2004). Congressional party defection in american history. Legislative Studies Quarterly, 29(4), 545–568.

    Article  Google Scholar 

  • Peltzman, S. (1984). Constituent interest and congressional voting. Journal of Law and Economics, 27, 181–200.

    Article  Google Scholar 

  • Poole, K. T. (1998). Recovering a basic space from a set of issue scales. American Journal of Political Science, 42, 954–993.

    Article  Google Scholar 

  • Poole, K. T. (2005). Spatial models of parliamentary voting. Cambridge: Cambridge University Press.

    Book  Google Scholar 

  • Poole, K. T. (2007). Changing minds? Not in congress!. Public Choice, 131(3), 435–451.

    Article  Google Scholar 

  • Poole, K. T., & Rosenthal, H. L. (1985). A spatial model for legislative roll call analysis. American Journal of Political Science, 29, 357–384.

    Article  Google Scholar 

  • Poole, K. T., & Rosenthal, H. L. (1991). Patterns of congressional voting. American Journal of Political Science, 35(1), 228–278.

    Article  Google Scholar 

  • Poole, K. T., & Rosenthal, H. L. (2000). Congress: A political-economic history of roll call voting. Oxford: Oxford University Press on Demand.

    Google Scholar 

  • Poole, K. T., & Rosenthal, H. L. (2000). Voteview manual. Pittsburgh: Carnegie Mellon University.

    Google Scholar 

  • Poole, K. T., & Rosenthal, H. L. (2017). Congress: A political-economic history of roll call voting. Oxford: Oxford University.

    Google Scholar 

  • Python Core Team. (2018). Python: A dynamic, open source programming language.

  • R Core Team. (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.

  • Rstudio Inc. (2018). R Markdown.

  • The Library of Congress. (2017). American memory. Journals of Congress. Accessed January 22, 2017.

  • United States House and Senate Historical Office. (2017). The biographical directory of the united states congress. Accessed 22 January, 2017

Download references


This project was made possible by support from The William and Flora Hewlett Foundation (Grant #2016-3870), the National Science Foundation (NSF-SBS-0611974), University of California Los Angeles Social Science Computing, and the University of Georgia.

Author information

Authors and Affiliations


Corresponding author

Correspondence to Jeffrey B. Lewis.

Rights and permissions

Reprints and Permissions

About this article

Verify currency and authenticity via CrossMark

Cite this article

Boche, A., Lewis, J.B., Rudkin, A. et al. The new preserving and continuing Keith Poole’s infrastructure for scholars, students and observers of Congress. Public Choice 176, 17–32 (2018).

Download citation

  • Received:

  • Accepted:

  • Published:

  • Issue Date:

  • DOI:


  • Roll call voting
  • United States Congress
  • Historical congressional data
  • Members of Congress
  • US Political Parties