The Squire Index measuring the professionalization levels of American state legislatures was created more than three decades ago (Squire Reference Squire1992). Scores for it were later extended back to 1979 and with this update for 2021 they cover a 42-year time span. In this note, I review some recent finding using the measure and explain how the 2021 update was compiled. I then examine the index’s stability over the years for which it has been measured, present two alternative measures to explore a problematic aspect of the 2021 index, and finally investigate how each of the index’s three components has moved relative to Congress between 1979 and 2021.
The impact of legislative professionalization
Since its inception, the index has come to be used as an explanatory variable in a wide range of studies examining both the way legislatures operate and the sorts of public policies that they (and state governments more generally) adopt (Squire and Moncrief Reference Squire and Moncrief2020, 63–64). In the last few years, for example, it has been demonstrated that professionalization impacts the way legislatures develop and process legislation. Less professionalized legislatures have been shown to be more likely to copy bills produced by other bodies than are more professionalized bodies (Jansa, Hansen, and Gray Reference Jansa, Hansen and Gray2019). Bureaucrats have greater involvement in crafting measures in less professional legislatures than they do in more professionalized ones (Kroeger Reference Kroeger2022). The propensity for a chamber to amend bills that originated in the other chamber goes up with professionalization (Makse Reference Makse2022). More professionalized legislatures experience lower concurrence rates in the second chamber for bills passed by the first chamber (Brown and Garlick Reference Brown and Garlick2023). The likelihood of a legislature using conference committees to resolve differences in legislation increases with professionalization (Emrich 2022). Laws produced by more professionalized legislatures are less likely to be vetoed by the governor and are given greater deference and less scrutiny by state courts of last resort (Armaly Reference Armaly2020; Ricknell Reference Ricknell2023).
Member behavior changes as well. Legislators in more professionalized legislatures are more active users of Twitter (Payson et al. Reference Payson, Casas, Nagler, Bonneau and Tucker2022). They are more likely to run for higher office and to win because of their greater resources (McCrain and O’Connell Reference McCrain and O’Connell2023). The way lawmakers interact with their constituents differs. Lawmakers in more professionalized bodies are more responsive to constituent emails, and they are less likely to discriminate against requests from racial minorities or the undocumented (Garcia and Sadhwani Reference Garcia and Sadhwani2023; Landgrave and Weller Reference Landgrave and Weller2020). Impacts are also seen in the way members interact with each other. There is, for instance, less within gender collaboration on bills in more professionalized legislatures (Swift and Vander Molen Reference Swift and Vander Molen2021).
The increased analytical capacity produced by professionalization translates into different policy choices. Lawmakers in more professionalized bodies introduce more bills related to scientific policy issues (Bromley-Trujillo and Karch Reference Bromley-Trujillo and Karch2021). More professionalized legislatures are less likely to preempt affordable housing policies at the local level (Goodman and Hatch Reference Goodman and Hatch2023). The inclination to adopt more complex regulatory policies increases with professionalization (Nattinger and Kaskie Reference Nattinger and Kaskie2017). They also make greater investments in water quality programs and projects (Williamson, Morris, and Fisk Reference Williamson, Morris and Fisk2021). But not all relationships are necessarily positive for the institution or state. For instance, governments with more professional legislatures rely on budget gimmicks to a greater degree (Hendrick and Hu Reference Hendrick and Hu2020).
Compiling the index for 2021
The index for 2021 was compiled in the same fashion as its predecessors (Squire Reference Squire1992; Reference Squire2000; Reference Squire2007; Reference Squire2012; Reference Squire2017). Legislative compensation figures for 2021 were taken from data provided by the National Conference of State Legislatures (NCSL). Most states now pay an annual salary and that figure is used in the index. Salary figures for the states that still pay a per diem or a weekly wage were estimated as previously, calculating them based on the length of the regular legislative session. Staff numbers were taken from the 2021 survey of legislative staff conducted by the NCSL. As before, the “total staff during session” figure, which includes both permanent and session-only staff, was used. The staffing numbers gathered by NCSL are the most authoritative available. The fact that they are only collected roughly every six years accounts for the gaps in the years measured by the index.
As with the 2015 index, the most problematic component in the 2021 index is measuring days in session, which was calculated as usual by averaging the two most recent session years for each state, in this case, the 2019 and 2020 sessions. The data were taken from the appropriate editions of the Book of the States. A few states reported legislative days for those years, some gave calendar days, and the rest only provided the dates that their regular sessions convened and adjourned. Calendar days were converted to legislative days using the same five-sevenths calculation that was employed for earlier scores. Given the unusual circumstances under which legislatures operated in 2020 because of the COVID pandemic, more states than typical waited to adjourn their sessions until later in the year. Consequently, the number of days in session calculated for those states may be inflated relative to their norm. It should be kept in mind that even though those states did not necessarily meet regularly in floor sessions throughout the year, leaving their sessions open longer meant that they still commanded their members’ time.
As with the earlier indices, each of the components for the 2021 measure was compared to the comparable component for Congress. Each of the components was equally weighted and a final overall score calculated. A score of 1.00 suggests complete equality with Congress on professionalization; a score of 0.00 suggests complete inequality. The Appendix provides information on all of the data sources used in compiling the 2021 index.
The 2021 legislative professionalization index: Stability and change
The 2021 legislative professionalization scores are presented in Table 1, along with the scores for the previous six years for which there are measures. The 2021 index contains no surprises and most states held roughly the same position as they had in earlier years. The five most professionalized state legislatures were again all large population states: California, Pennsylvania, New York, Massachusetts, and Michigan. The five least professionalized state legislatures were once more all small-population states: New Hampshire, Wyoming, North Dakota, Montana, and Utah. The notion that relatively little changes in the scores from measurement to measurement is confirmed by the correlations presented at the bottom of Table 1. The 2021 score correlates with the 1979 score at an impressive .830. And, as we might expect, the correlation with the 2021 score gets stronger with each successive index. Legislative professionalization levels as measured by this index are remarkably stable. This should not be a surprise because institutions generally do not change dramatically from year to year, or even from decade to decade.
Sources: The scores for 1979 to 2009 are from Squire (Reference Squire2012, 308–309). The score for 2015 is from Squire (Reference Squire2017).
That is not to say that nothing changes with state legislatures. Between the 2015 and 2021 scores, for example, New York lawmakers enjoyed a substantial boost in their listed salary, as Alabama lawmakers had between the 2009 and 2015 scores. Staff numbers bounce around only a little, but some of the states with the largest staffs saw their numbers decline between recent measurements. During the first two decades of this century, both Arkansas and Oregon shifted to annual sessions from biennial sessions, increasing their days in session scores; similar swings for the same reason had occurred for several other states between scores in earlier years. The index, of course, picks up these fluctuations, but because it has three equally weighted components, unless all three components for a state shift noticeably in the same direction, fluctuations in the overall scores are likely to be muted.
A potential concern with the 2021 professionalization measure
Arguably, the biggest potential reservation with the index is the way days in session are measured. As noted in the discussion of the construction of the 2015 index, that particular component is apt to be noisy, at least for some states (Squire Reference Squire2017, 363–66). And, as mentioned above, the impact of the pandemic on legislative sessions in 2020 may have exacerbated that problem in the 2021 index. Alternative approaches to constructing the 2021 index along with state rankings are offered in Table 2 as a way to better understand the potential impact of this problem.
One way to assess it is to continue to calculate days in session as it has always been done, but to average them over the preceding four years rather than the preceding two years. That approach should reduce the impact of any aberrant session. It may make sense, not just for diluting the possible impact of 2020, but for attenuating the impact of any single year which, for any of a variety of reasons, a legislature met for more or fewer days than it typically does. The middle columns of Table 2 present a revised 2021 index incorporating days in session as the average number between 2017 and 2020 rather than just between 2019 and 2020. The additional data were again gathered from appropriate editions of the Book of the States.
This version of the 2021 index correlates with the traditional version at .988, indicating that using two additional years changes little. Looking at the scores for the 50 states, however, seven cases are noticeably different. In two cases, Maine and New Hampshire, the legislature reported having met for many more days in session in 2017 and 2018 than they reported for 2019 and 2020. Accordingly, their scores in the alternate 2021 measure are higher than in the traditional measure. The other five cases are states where one year either had many more days than did the other three years (Vermont and Wisconsin reported many more days in 2020) or far fewer days (Michigan and Pennsylvania reported far fewer days in 2017, as did New York in 2019). Those differences had a conspicuous impact on their scores. Thus, averaging days in session over four years rather than two years does not eliminate noise in that component.
The last two columns in Table 2 present another alternative version of the 2021 index, this time dropping the days in session component altogether and reporting only the scores using the other two components. This version still correlates with the traditional index at a high level: .812. But the mean and median scores are deflated, as are the scores for some states. This suggests that, in addition to being important for theoretical reasons, days in session also play a significant role in moderating the professionalization gap between state legislatures and Congress. But, given the strong positive correlation between this alternative measure and the traditional measure, it might be the case that substituting one for the other would not produce substantively meaningful differences in the statistical impact of legislative professionalization in many analyses.
Moreover, it should be noted that dropping days in session would effectively treat state legislatures as having met for the same number of days. We know that is not true and that the number of days in session varies greatly across the states. We also think that the amount of time devoted to policy development is an important contributor to legislative capacity. Thus, even with the noise inherent in the traditional way days in session are measured, including it in the index is theoretically and statistically preferable to omitting it.
Are state legislatures getting more professional or is Congress just stagnating?
Mean and median scores for each of the years the index has been measured in the traditional manner appear at the bottom of Table 1. The 2015 and 2021 indices suggest that state legislatures are, on average, becoming slightly more like Congress. But it might be argued that Congress is becoming marginally more like the average state legislature. The reason is that in recent years Congress has stagnated in regard to the three professionalization components. Members of Congress have not received a pay increase since 2009. Both houses sometimes meet for fewer days than they did in past decades. Congressional staff levels have changed little. Overall, these developments have allowed state legislatures to faintly close the gap as some of them have, at least marginally, raised member pay, met for more days in session, or increased staff.
To further examine how state legislatures changed relative to Congress, median state legislative scores for each of the three components at each time the index has been compiled are presented in Table 3. These numbers underscore an important point to keep in mind about the index: Both sides of the calculations are moving targets. The components for state legislatures change at least a little and so do those for Congress.
Sources: The median professionalization score is from Table 1. The other medians were calculated from data gathered from sources from 1979 to 2009 given in Squire (Reference Squire2012, 308–309), and for 2015 in Squire (Reference Squire2017).
Over time, the state legislative components largely behave as we might expect relative to Congress. Median state legislative pay held relatively steady through 2009, and then increased slightly as Congress failed to raise its pay. The median state legislative staff per member figure increased between 1979 and 1986 and then fluctuated trivially over the next couple of decades. The median state legislative days in session figure documents that, over the years the index has been calculated, state legislatures were more like Congress on that component than they were on the other two components. But the median days in session figure is, as anticipated, more volatile. But its volatility needs to be understood in context; it was volatile for Congress and some state legislatures, but not for many other state legislatures which stuck close to the same number of days in session year after year. Compared with Congress the median state legislature met for relatively more days in session 1979 and 1986 and then again in 2015 and 2021, but for relatively fewer days in 1996, 2003, and 2009. Thus, while combining the three components of professionalization makes sense theoretically, it is also important because it works to mitigate the impact of swings in any single component. As a result, over time, the index stays stable.
It should be kept in mind as well that even with the stagnation of Congress on the three components, there is still a wide gap between it and even the most professionalized state legislature. Members of Congress still receive much higher pay than the highest-paid state legislature and they still enjoy far more staff than even their best-resourced state-level counterpart. And, of course, Congress continues to be a year-round institution that is in session for at least a few weeks almost every month. That means that it still establishes the baseline against which to assess the capacity of state legislatures.
Discussion
The updated 2021 scores provide a seventh data point measuring legislative professionalization calculated in the same fashion using the same kinds of data. The time span covered by these measures now stretches to 42 years. Given the measure’s characteristics, it would appear to be a valuable explanatory variable for analyses of both institutional and policy changes across the American states. And, of course, the 2021 scores should be useful in helping to explain recent legislative behavior and policymaking.
The measure is easily incorporated as an independent variable in cross-sectional analyses and that is the way it is commonly employed. Over time analyses could conceivably be more complicated because of the movement on both sides of the way the measure is calculated. One proposal to address those concerns was to “anchor the professionalism standard set by Congress … by calculating the grand mean for Congress” (Squire Reference Squire2007, 222–24). That approach does not appear to have gained traction. A more popular alternative was proposed by Bowen and Greene (Reference Bowen and Greene2014). They calculated yearly scores, using annual salary and session data, and substituting legislative expenditure figures for staffing numbers because of the latter’s infrequency of collection. The limitation of that approach is that it is not always clear what is incorporated into legislative expenditures and how directly it translates into staff resources. Measures using it, however, correlate reasonably well with the traditional index.
More important for consideration of this question is a reality that has been noted at several points in the above discussion: The components of legislative professionalization do not change much from year to year. Indeed, they exhibit considerable stickiness. In 2021, for example, legislative salaries had not been changed at all for more than a decade in 14 states, in eight of those states pay had not changed since the latter part of the twentieth century and in one state – New Hampshire – it had not changed since the nineteenth century (1889). Only 11 states do not place any limits on how many days their legislatures may meet in session, allowing their numbers to potentially fluctuate noticeably. Strict limitations are set by state constitutions in 28 states and statutes, legislative rules, or restrictions on the number of days per diems may be paid constrain meeting days in the rest of the states (Squire and Moncrief Reference Squire and Moncrief2020, 75). As a result, in most states, the number of days in session will only change slightly from one year to the next. Staffing numbers tend to shift up and down only slowly over time. All of this suggests that there will be little volatility in the professionalization components. That stability allows the Squire index to be of use in a wide range of studies.
Data availability statement
Replication materials are available on SPPQ Dataverse at https://doi.org/10.15139/S3/2VORBN (Squire Reference Squire2023).
Funding statement
The author received no financial support for the research, authorship, and/or publication of this article.
Competing interest
The author declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Author Biography
Peverill Squire is a Professor of Political Science at the Truman School of Government and Public Affairs at the University of Missouri and holds the Hicks and Martha Griffiths Chair in American Political Institutions. Among his recent books are Reforming Legislatures: American Voters and State Ballot Measures, 1792–2020 (University of Missouri Press, 2024), The Right of Instruction and Representation in American Legislatures, 1778 to 1900 (University of Michigan Press, 2021), and The Rise of the Representative: Lawmakers and Constituents in Colonial America (University of Michigan Press, 2017).
A. Appendix: Data Sources
Congressional days in session for 2019–2020 were calculated by the author using data from, https://www.senate.gov/legislative/ResumesofCongressionalActivity1947present.htm.
Congressional pay was taken from Congressional Research Service, “Congressional Salaries and Allowances: In Brief,” December 16, 2022.
Congressional staff per member was calculated by the author using data from, https://www.brookings.edu/multi-chapter-report/vital-statistics-on-congress/, table 5.1.
State legislative days in session were calculated by the author using data from Book of the States 2018, Book of the States 2019, Book of the States 2020, and Book of the States 2021.
State legislative pay was calculated by the author using data from NCSL, https://www.ncsl.org/research/about-state-legislatures/2021-legislator-compensation.aspx.
State legislative professionalization scores for 1979 to 2009 were taken from Squire (Reference Squire2012, 308–309). The scores for 2015 were taken from Squire (Reference Squire2017).
State legislative staff data were taken from NCSL, https://www.ncsl.org/research/about-state-legislatures/staff-change-chart-1979-1988-1996-2003-2009.aspx.