75 Years of Institutional Change in State Legislatures

The 75th anniversary of The Book of the States inspires a look back at the first 1935 edition to determine what we can learn about state legislatures in the 1930s compared to the 2000s, and to recount the history of change in these institutions. Most of the story of change relates to the strengthening of legislatures—the process by which they have increased their ability to make decisions independently of the executive branch and lobbyists.

About the Author

Karl T. Kurtz is director of the Trust for Representative Democracy at the National Conference of State Legislatures. He wasn’t around for the 1935 edition of The Book of the States, but has worked for The Council of State Governments or NCSL for more than half of the 75-year span of this publication. He last wrote the state legislatures chapter for The Book of the States for the 1974-75 edition.

Book of the States, Chapter 3

Download the PDF Version of this Article

  Table A: State Legislative Membership and District Populations, 1933 and 2009 (Population in Thousands): Download in Excel or PDF

Table B: Length of Legislative Terms of Office: Download in Excel or PDF

Table C: Legislative Compensation Comparison: Download in Excel or PDF

Table D: Number of Standing Committees: Download in Excel or PDF

The three standard measures of the effort to increase legislative capacity, which political scientists call “professionalization,” are the time that legislators spend on the job, the compensation they receive and the staff resources available to them. The chapter on legislatures in the 1935 Book of the States has only nine tables, compared to three times as many in modern editions. Nonetheless, we can learn a lot about how dramatically different the legislatures of today are on the dimensions of time, compensation and staff compared to the era of the 1930s.


The time demands of legislative work have increased greatly since the 1930s. The most basic measure of this is the predominance of annual legislative sessions today compared to biennial sessions in the earlier period. In 1933, five states held annual sessions, 43 met biennially and one, Alabama, met only every four years. Today it is the reverse: Only five states continue to meet biennially and the remaining 45 have annual sessions. Most of this shift to annual sessions occurred during the heyday of legislative reform in the late 1960s and early 1970s. By 1974, 42 states had switched to annual sessions. Arkansas became the most recent to switch when voters approved a constitutional amendment in 2008.

The switch to annual sessions is the most obvious evidence of more time on the job by legislators today. Comparing the number of days spent in session is more difficult within the confines of the information in the Book of the States, but we can reach a few general conclusions. In addition to switching to annual sessions, about half the 31 states that had constitutional limits on session length in 1931 increased the number of days allowed in session or gave the legislature more flexibility in how to use those days. The other half of the states in this category kept their limits on session length but applied them to annual, as opposed to biennial, sessions. Virtually all the states that had no limit on session length in the 1930s spend more time in session today.

Special sessions also occupy the time of legislators. In 1933, 35 states held 43 special sessions, while in 2009, 23 states held 63 special sessions. Presumably, the reduction in the number of states holding special sessions is due to the growth in annual sessions, making it less necessary for governors or legislative leaders to convene the legislature extraordinarily.

Requests for service from constituents also affect the time demands on state legislators. In general, the larger the population of a legislative district, the more demand there is for help in dealing with constituents’ problems. The population of legislative districts in each state is determined both by the number of legislators and the state’s population.

In 1933, 7,527 state legislators served a 48-state population of 125 million people. Seventy-five years later, the U.S. population has increased to 306 million, but the total number of legislators has declined to 7,382, even after adding two more states to the union. That’s a net reduction of 145 state legislators.

But this reduction in the total number masks differences between houses of representatives and senates. In the house chambers, two factors caused a net reduction of 295 legislators. First was Nebraska’s conversion to unicameralism in 1937, which eliminated 100 house members. Second, a prevailing belief of the legislative reform movement of the 1970s that smaller legislatures were more efficient and effective led Connecticut, Massachusetts and Vermont to eliminate more than 80 house seats each, followed by Georgia, Illinois, Ohio and Rhode Island, each of which also reduced the number of their house seats by 20 or more. The total number of senators, on the other hand, actually increased by 150 during this period. The expansion of senates in Maryland, New Jersey, New Mexico and New York, along with the addition of Alaska and Hawaii as states, account for three-quarters of the additional senators nationally.

That state legislators today represent far more people than they did 75 years ago is perhaps an obvious point given the slight decline in the number of legislators and a large increase in population. When we look at individual state data on the ratio between state population and number of members in each chamber, though, some of the differences are dramatic. In Nevada today, House members on average represent 26 times as many people as they did in 1933. Florida’s ratio between population and House members is 16 times greater, and California’s has increased tenfold. Changes in ratios of senators to population are somewhat less dramatic, but nonetheless large in many states.

The only states in which legislators today represent almost the same number of people as they did in 1933 are North Dakota, whose population has declined slightly, and West Virginia, whose population has increased slightly, as has the number of legislators.

Closely related to the time demands of legislative work is the length of legislative terms. The need to campaign for elections greatly affects the work of legislators. The longer the term of office, the less time lawmakers need to spend on campaigns.

The only significant change in the length of terms in houses of representatives is the switch from one-year to two-year terms in New Jersey and New York. In senates, though, most states have changed from two- to four-year terms, thereby demanding less time for campaigns.

As an aside to this discussion of time demands of legislative work, in 1933 there were no term limits for state legislators. Successful voter efforts to impose term limits on legislatures in 15 states through the initiative process did not begin until the 1990s. Term limits have created a new management problem, unforeseen in the 1930s, for legislatures in the 2000s.


To compensate for longer sessions, the much larger number of constituents and the increases in responsibility, the pay for legislators has changed dramatically since the 1930s. In 1935, legislators in 19 states were paid less than $300 a year, with Kansas the lowest at $75 per year and New Hampshire close behind at $100. In 13 states, legislators received $400 to $600 in annual pay. Only nine states paid their lawmakers $1,000 or more for legislative service, led by New York at $2,500. For the most part, the same larger population states that pay their legislators more today paid more in 1935.

If we restate those 1935 salaries in 2007 terms, to take into account the effect of inflation, they are not comparable to today. Legislators in 1935 were paid substantially less compared to today. Lawmakers in 31 states received less than $7,000 per year in today’s dollars. Currently, only New Hampshire pays its legislators $100 per year. And in 1935 nine states paid their legislators more than $15,000 (in 2007 dollars), compared to 40 states today.

Every state has increased real compensation for legislators except New Hampshire. It is two to five times higher in most states and as much as 10 to 20 times higher in a few states. Keep in mind that 1935 was at the end of the Great Depression and that similar increases in real compensation occurred for American workers in most industries and professions since that time. In fact, the increases in compensation that legislators have received in most states appear to be due more to overall improvement in personal income than to the professionalization of legislatures. The largest increases in legislative pay occurred in most states before 1972 (the earliest year for which I collected these data), which is early in the period of legislative reform. Legislator pay has declined in real terms in 20 states since 1972, and in another seven states the inflation-adjusted pay increases have been less than 10 percent in 35 years.


The 1935 edition of the Book of the States provides limited information about legislative staff. After a lengthy description of the history and work of the Wisconsin Legislative Reference Library, which was formed in 1901 and is generally regarded as the first “legislative reference service,” a chapter in the book lists 22 states that had established research offices by 1935 and provides this sketchy description of the numbers of staff:

Many of the bureaus have developed into effective organizations, although some of them have only one or two staff members. Among those which have the desirable combination of high standards of workmanship and staffs of fairly adequate size are California, Connecticut, Illinois, Indiana, Massachusetts, Maryland, New York, Pennsylvania and Wisconsin. Each of five states—California, Maryland, New York, Pennsylvania and Wisconsin—has eight or more people engaged in legislative reference work on a full-time basis.

If we make a wild guess that the five states with “adequate size” had 10 staff on average, and estimate that the other 17 states with reference bureaus averaged, say, five people each, then give the remaining 26 states that are not listed as having research offices two staff each, that would mean that there were approximately 200 to 250 full-time staff working for the 48 state legislatures in the mid-1930s.

This rough estimate may leave out other full-time legislative staff, but probably not a lot. The legislative clerks and secretaries, whose offices long pre-date the development of research services, were staffed in those days almost entirely on a part-time basis. The bill drafters and legal counsels, fiscal analysts, legislative auditors, committee support, security personnel, information technology experts, and personal and partisan staff that populate today’s state legislatures simply did not exist in the 1930s, at least not full-time. To the extent that legislatures used such specialists, they borrowed them from executive agencies or state universities or hired them for the session only. More importantly, before the advent of professional staff in legislatures, the members relied on themselves and their fellow members for expertise on issues.

Even with this roughest of estimates, it seems safe to say that the total number of full-time staff working for state legislatures was less than 500 in 1935. Today, according to the National Conference of State Legislatures’ last census in 2009, there are approximately 28,000 full-time staff who work for state legislatures. Sixteen states by themselves have staffs of more than 500 each.

By the 1960s every state legislature had some kind of legislative research service, most of them working in a central, nonpartisan staff office. In the late 1960s and early 1970s, the majority of states established legislative fiscal offices to provide independent evaluations of the budget prepared by the executive branch. In the 1970s, many states transferred state auditors’ functions to the legislative branch and established program review offices, either on their own or as part of a legislative auditor’s office. During this period, clerks and secretaries’ offices and security services became full-time, professional operations in most states. Some states split their research and committee staff functions to serve the house and senate independently. The most populous states added personal staff to members. Between 1968 and 1974 legislative staff more than doubled.

In the 1980s, states added information technology staff and established legislative information offices. Partisan leadership, caucus and members’ personal staff continued to grow throughout the 1980s and 1990s to the extent that today, they make up slightly more than half of all legislative staff.

By the late 1990s, though, the growth of legislative staff had leveled off, and in the 2000s there has actually been a slight decline in the number of legislative staff.

Other Legislative Organization and Procedure

The nine tables in the 1935 Book of the States reveal a few other things about the legislatures of that era compared to today. The one table that has barely changed at all is the list of official names of legislative bodies. The only changes from 1933 to today are that Montana changed its official name to Legislature from Legislative Assembly, and Nebraska eliminated its House of Representatives and became a unicameral body. Otherwise, this table is identical today except for the addition of information about the legislatures of the District of Columbia and the territories and commonwealths of American Samoa, Guam, the Northern Mariana Islands, Puerto Rico and the U.S. Virgin Islands. (The information about the U.S. territories and commonwealths that is common to modern editions of The Book of the States is missing from the early volumes.)

One of the principle strategies of the legislative strengthening movement of the late 1960s and early 1970s was to urge legislatures to reduce the number of legislative committees, rationalize their structure and jurisdictions, and make them effective gatekeepers in screening legislation for the chamber as a whole. The need for this reform is evident in data from the 1935 Book of the States on the number of standing committees in each chamber. Well over half the senates and houses had more than 30 standing committees each, and very few had 20 or fewer. Today’s numbers reflect the fact that those recommendations for committee restructuring were widely adopted in the states: The range of 11 to 20 standing committees has become the norm for state senates, and four out of five houses of representatives now have 30 or fewer committees.

Another plank in the legislative strengthening platform was establishing detailed and meaningful deadlines for processing legislation to avoid end of session logjams. Both the first and the most recent editions of the Book of the States contain tables on deadlines for the introduction of bills, but the information in these tables only scratches the surface of the topic and does not allow us to reach any meaningful conclusions about deadlines.

Another table in the first volume of the Book of the States reveals that lieutenant governors had the power to appoint the standing committees in 16 of the 48 states. Today that is true in only two states as state senates have moved to place management responsibility for their chambers in the hands of senators whom they elect to leadership positions rather than in the statewide elective office of lieutenant governor.

Unfortunately, this tidbit of information about legislative leadership powers is the only thing we can draw on from the 1935 edition about how legislative leadership has changed over time. There are numerous other legislative strengthening topics we cannot learn about from the nine tables of that otherwise admirable first effort at reporting on the organization and procedure of America’s legislatures. Topics that have become routine in recent editions of the Book of the States—benefits for legislators, pre-filing and carryover of bills, vetoes and veto overrides, the appropriations process and requirements for fiscal notes, the number of bill introductions and enactments, and procedures for standing committees—were not addressed in volume one.

On the other hand, there are some intriguing miscellaneous facts about legislatures in the first edition. For example, we learn that in 1933, 35 percent of all members of Congress previously served in state legislatures. For the last three decades of American politics, that proportion is substantially higher, hovering consistently around 50 percent.

The 1935 edition also contains the names of all 7,500 state legislators in the United States. By comparing those names to the same list in the next (1937) edition, we can estimate the rate of turnover in the membership of legislatures in those days. Nationally the average proportion of newly elected members serving in the houses of representatives was 55 percent, which compares to a turnover rate of about 25 percent in legislatures in the 2000s. This dramatic change in turnover is a result of the legislative strengthening changes in time, compensation and staff previously described. As the capacity of legislatures has increased, so too has the desirability of the job of legislator, thereby reducing the number who voluntarily step down from office. As the job has become more attractive, incumbent lawmakers have made skillful use of the resources of experience, information and staff available to them to ensure that they are re-elected.


The transformation in the institutional capacity of state legislatures that is summarized here coincided with resurgence in the policy roles of legislatures in the federal system. The legislatures of the early 1930s were emerging from a period in which the U.S. Supreme Court had consistently ruled that state government had little role in the regulation of private business, and they were entering into the great expansion of federal power and authority brought about by the New Deal and World War II. Legislatures of the period were notoriously malapportioned with overrepresentation of rural and farming interests at the expense of urban, industrial populations.

The Supreme Court’s one person, one vote decisions of the 1960s provided an initial spark to the legislative strengthening movement by forcing legislatures to be more representative of state population as a whole and bringing in a new generation of legislators. State legislatures remained quiescent in their policy roles until the 1970s and 1980s. (The Sometime Governments is the title of the most important book urging legislative improvement and reform in 1970.) Then the “new federalism” policies of the Nixon and Reagan administrations turned responsibility back to the states, and in the 2000s, states took on even more policymaking responsibility when federal domestic programs were severely cut due to economic conditions. The states and their governors and legislatures had no choice but to assert their authority and responsibility. Thanks to the legislative strengthening movement, they had the capacity to do so.

The challenges legislatures face today are not the same as in 1935. At least for the time being, legislatures have built their capacity and, except for isolated cases, are unlikely any time soon to be adding staff and facilities, increasing their compensation or expanding their time on the job. New challenges that face the modernized American legislature include finding the resources and will to overcome a severe economic downturn, coping with term limits, mitigating the effects of intense partisan conflict, controlling the effects of money in politics, harnessing the power of information technology, improving methods of redrawing their legislative districts, and dealing with intense public distrust and cynicism toward democratic institutions.