by Karl Kurtz
Continuing yesterday's subject of historical archives, a remarkable book has recently come across my desk: Party Affiliations in the State Legislatures: A Year by Year Summary, 1796-2006 by Michael J. Dubin. It's a state by state listing of the party lineups for every year since partisan contests for the legislature began and party affiliations were listed.
According to the cover jacket, the author is a retired school teacher who lives in Arizona and has published similar volumes of historical election data for governors, presidents and Congress.
Don't look for any analysis in this book. There are only two summary tables, one listing the total number of legislatures or chambers under the control of each political party since 1834, and the other showing the length of legislative terms in each chamber and state since 1790. What you get for each state is one paragraph each on the history of statehood and constitutions, terms of elections, legislative districts, size of membership, redistricting, and dates of elections, followed by the listing of party composition after each election year.
For example, in my native state of Ohio, I could learn that in 1827, the first year for which party affiliations were listed in the state, the Senate was controlled 22-13 by the National Republicans over the Jacksonians, and the House was 44-28 for the same two parties. In 1901, when favorite son Republican William McKinley was President, Republicans had a 21-12 margin over Democrats in the Senate and of 68-42 in the House. After the election of 1958 (the first one that I can remember), when there was a right to work measure on the ballot in Ohio that mobilized unions to get out the vote, the Democrats swept into power with majorities of 20-13 and 78-61 in the Senate and House respectively.
Having been sent into the stacks in the Library of Congress nearly 40 years ago to collect some of the very same data for my graduate school adviser, Walter Dean Burnham, I know how elusive this information is, especially in the pre-Civil War period. You find it in very obscure publications or newspapers, and once you find it you're faced with a confusing array of party names and affiliations that vary greatly from state to state. Interestingly, Dubin makes no mention of Burnham's dataset of the same materials contained in the University of Michigan's Inter-University Consortium for Political and Social Research. Dubin appears to be unaware of this resource, which contains party composition of legislatures, 1834-1985.
I spot-checked two states (four chambers) in two different years in the 19th century in Dubin's book against Burnham's database and found significant differences in the numbers of legislators listed for each party, apparently having to do with how "other" parties were handled. In the eight cases that I checked, Dubin's and Burnham's numbers agreed in three instances and disagreed in five. I had no way of knowing which one was more accurate.
Nonetheless, Dubin's book is a huge and painstaking achievement that will be a valuable tool for historical researchers on state legislatures and legislative elections. At NCSL we're pleased that we are his source for election results data since 1998.
Dubin's summary table of the length of legislative terms provides some interesting historical perspective. As I have written before, in the early years of American history, legislative terms were substantially shorter than they are today. Dubin's table shows that in 1800, for example, 13 of the 16 (81 percent) houses of representatives had one-year terms, two had six-month terms (!) and only one had a two-year term. Similarly, 53 percent of the senates had terms of two years or less, 13 percent had three-year terms and only 33 percent had four-year terms.
This pattern changed slowly in the early part of the 19th century but more dramatically in the revisions of many state constitutions in the 1840s and 1850s and the addition of new states in that period. By 1880, after the Reconstruction rewriting of constitutions in the South, the six-month house terms had disappeared, only five houses had one-year terms, and 32 of 38 (84 percent) had two-year terms. In senates in 1880, two had one-year terms, 11 had two-year terms, one had a three-year term, and 24 (63 percent) had four-year terms.
Today 45 of 49 states have two-year house terms, and four have four-year terms. Among senates only 13 states have two-year terms and 37 have four-year terms. Dubin's history of terms of office doesn't say anything about the 15 states that have term limits today.
Here are charts of the history of the length of legislative terms at the beginning of each decade. (Senates on left, houses on right. Click to view full size.)