How Autonomous are Your Legislative Committees?

by Karl Kurtz

The most interesting thing to me about Nancy Martorano's prize-winning article on legislative committee systems was something that was not in the published version of the article: the actual autonomy scores for each state's committee systems.  I thought that these scores might be of interest to legislative junkies, so I contacted Nancy about it. She provided me with committee system autonomy scores for both legislative chambers in all 50 states for 2001-02 and gave me permission to publish them here.

These scores are based on the formal procedures--constitutions, statutes and rules--that affect the ability of committees to function as autonomous actors in the legislative process.  The scores are a measure of the freedom of committee action on 11 different legislative procedures.  On each of these 11 procedures Martorano gave a score ranging from -1 to 1 to all 99 legislative chambers.  For example, she judged that legislative committees that can kill bills (+1) are more autonomous than ones that are required to report all bills to the floor (-1).  Or, committees that can offer substitute bills in lieu of those referred to them (+1) are less constrained than ones in which the rules prohibit them from substituting bills (-1).  No mention of these subjects in the rules of procedure scored 0.  For a complete list of the procedures, read below the jump.

Before presenting the scores, let me hasten to say that committee autonomy does not necessarily mean committee strength.  The autonomy measure deals only with formal procedures.  But it's the informal procedures--the norms and customs--of a chamber, combined with the formal rules, that determine the strength of committees.  For example, rules of procedure may make it easy for the chamber to remove a bill from committee or to amend its recommendations on the floor, but in practice the chamber may never take such actions and may routinely follow committee recommendations.  Unfortunately, the customs and practices that are critical to determining the strength of committee systems are very difficult to measure across states.   

Nonetheless, the autonomy scores are interesting.  Theoretically, the autonomy scores can range from -11 to 11.  In practice, in houses of representatives in 2001-02 the range of scores was from .5 (Kansas) to 10 (Tennessee), and in senates it was from -1 (New Hampshire) to 11 (Illinois and Michigan).  The following maps present the approximate scores for each chamber (senates on the left, houses on the right--click to enlarge).

There is no apparent explanation for the variation of these scores from state to state or even from chamber to chamber.  Neither region nor level of professionalization of the legislature appears to explain them. Senate scores are generally higher than house scores: the average senate score is 6.0, while for houses it is 5.1.  In one way this seems odd because normally I would expect that the larger, more unruly bodies would have more need for the gatekeeping roles of committees.  On the other hand, the smaller senates are traditionally more collegial, so perhaps they trust their committees more.

Thanks to Nancy Martorano for making these scores available. They first appeared in “State Legislatures and Committee Systems: Developing a Framework for Comparatively Studying the Institutional Power and Functioning of Committees in the Legislative Process,” Paper presented at the annual meeting of the Midwestern Political Science Association, Chicago, Illinois (April 2004).

Here are the 11 procedures that go into the autonomy scores:

1. Do all bills have to be referred to committee?
2. Do committees have to consider all referred legislation?
3. Are committees required to report all bills back to the chamber?
4. Do committees have to report bills back to the chamber within a specified time?
5. How difficult is it for the chamber to remove bills from committees?
6. Do committees have the power to sponsor and introduce legislation?
7. Are committee amendments incorporated into bills before consideration by the chamber (i.e., the chamber does not consider committee amendments separate from the bill)?
8. Can committees offer substitute bills in lieu of those referred to them?
9. How difficult is it for the chamber to overturn an adverse committee report?
10. Do bills go on the calendar in the order reported out of committee?
11. Are there restrictions on amending bills on the floor?

For details on the calculation of the scores, see Nancy Martorano, "Balancing Power: Committee System Autonomy and Legislative Organization," May 2006, Legislative Studies Quarterly.

Note on the map for houses of representatives: Don't be confused by Nebraska.  Nebraska is shown in white on this map because the Unicameral does not have a house of representatives.  But white is also the color for the lowest score on this scale.  I wasn't smart enough to figure out how to show Nebraska as "no data" or "unicameral" on this map.

Comments

[rich miller]

You gotta be kidding me. Illinois gets an 11?

Committee chairs and spokespersons are appointed by legislative leaders here, not elected. Staff reports to those same leaders, not chairs or spokespersons. Members can be easily replaced by leaders. Leaders decide which committees will hear which bills and control the posting of committee times and dates.

What a bogus study.

[Karl Kurtz]

Thanks for the comment, Rich. As I said in the posting, the committee autonomy scores reflect only formal authority, not how the process actually works or how strong or effective the committees are. For what it's worth, Chris Mooney and Tim Storey's case study of Illinois (a "control" state) for our term limits book (http://www.ncsl.org/jptl/casestudies/CaseContents.htm#il) agrees with you about the dominance of the leadership and the relative weakness of committees in the Illinois General Assembly. Your comment and their findings effectively point out the limits of the committee autonomy scores.

[rich miller]

Perhaps the "committee autonomy scores" should then be adjusted to reflect reality. What's the point of a scoring system that produces completely bogus results?

[rich miller]

Perhaps the "committee autonomy scores" should then be adjusted to reflect reality. What's the point of a scoring system that produces completely bogus results?

[Nancy Martorano]

Unfortunately, when attempting to undertake a large scale study of the operations of state legislatures, it is often impossible to capture all of the informal nuances at work in each of the 99 state legislative chambers. This would require that I rely largely on the accounts of insiders and observers, the quality and detail of which can vary quite a bit from state to state. Thus, to ensure that all of the state legislative chambers were treated equally by my measure, I chose to focus only on structures, rules and operations pertaining to committees that could be found in formal legislative documents such as the chamber rules of procedure, state statutes and the state constitution. I do not believe that my measure is "bogus," it simply captures a different aspect of committee operations from what you would prefer. Further, what I measure are formal rules of procedure – meaning they can be invoked at any time. In my article, I make no claim that my measure is anything more than a measure of “formal” autonomy, and I acknowledge that this may not actually translate into real autonomy in the legislative process. What you describe in Illinois is a system where the members have informally chosen to allow the leadership a great amount of power over the legislative process. What my measure indicates about the Illinois state legislature is that if the rank-and-file members ever tire of this arrangement they possess quite a few formal tools via the committee system that they could employ to take power away from the leadership if they so choose.

[Rich Miller]

In other words, reality does not matter. Thanks.

[Todd Drafall]

I don't even know what to add to Rich's comments,
How about the fact that a Legislative leader in Illinois can replace a member of a committee for a vote when needed? or the use of the rules committee to keep what ever bills leadership wishes from ever seeing the light of day. Or scheudling conflicting committees to shift the balance. Or never calling the bill to the floor for a vote.
This is like saying Germany in the 30's was technically a democracy, or Cuba, or Russia. Sure the structure was there, but the effect and the reality of the situation simply doesn't hold. I'm not saying Illinois is a dictatorship, I am comparing to what might exist on paper and what Rich has already documented are two very different things and if you are going to develop a study, lets give some credience to financial independence of members and some test of the system against current practice. I suspect there are other lower scoring systems that are much more democratic and open then the high scoring Illinois system that is so very controlled

[ArchPundit]

The problem is that the formal autonomy of the Illinois Lege is exactly what gives leadership the power. By being able to kill a bill in committee, the leaders use that institutional rule to their advantage. It's only one example, but if you want to kill bills, referring them to committees that rely upon the leader for both continued time on the committee and campaign cash means the the leadership exerts exactly the kind of control Cox and McCubbins argue.

It strikes me that adding campaign finance as a variable might deliver very different results for you--and one that may well lead to partisan theory far more influential. In addition, operationalizing further mechanisms for committee choices might make a difference, though my guess is the big issue is campaign finance methods. In Illinois, most Lege members rely on leadership and so leadership is able to use that as the most basic tool in controlling members. And they do.

As you continue your work, I'd strongly suggest looking into that question.

[MF]

Having worked on legislative staff in Illinois, all I can say is: I want some of what you're smoking!! Leaders would regularly change the "membership" of the committee to protect legislators from tough votes. Invidivual members would have a "conflict" and be "replaced" by a "proxy". Nothing moves in the state without the blessing of leadership. Example #2315 of how the political science academic profession knows nothing about the subject they study.

[random political scientist]

"Example #2315 of how the political science academic profession knows nothing about the subject they study."

Not all political scientists are sold on this methodology. For my two cents, informal power trumps formal rules just about every time, though informal power is a little hard to measure.

One interesting alternative is presented by Richard Clucas at Portland State, who (if I remember the study correctly), interviewed legislators and staff in most of the states about (among other things) who they thought the most powerful players in the legislature were, and then used these responses to come up with a measure of legislative leader power. Illinois scores a little more predictably here. It's in the Spring 2007 issue of State Politics and Policy Quarterly (http://sppq.press.uiuc.edu/7/1/abstract1.html), which unfortunately and not very explicably is subscription-only.

Worth a read if you can find a copy. Don't give up on us just yet!

[Chris Mooney]

Although I try hard to avoid the "political scientists don't know jack about how politics REALLY work" argument, let me make three quick points here.
First, the 2007 Clucas article in SPPQ does use surveys of insiders to assess leadership power and, not surprisingly, finds the IL speaker rates second only to the NY speaker in 1995 (although he ranks 22nd in 1981-- just post-Cutback Amendment, again would be expected by those who know IL politics, I think).
Second, those who study comparative state politics understand the dangers of random error only too well. That is what we are seeing in the Martorano measure with the IL data point. It is like the voter who is white, male, 52, makes $150K, watches NASCAR, and is a veteran-- but votes Democratic. Maybe George Bush used to beat him up on the playground when they were kids-- who knows. But just because we got that guy wrong doesn't mean that generalizations made about voters overall don't get us some way toward understanding the political world a little better. The problems with studying the states (among many) are a) we only have 50 (or 99 or 49 or whatever) cases, so random error is especially a problem, and b) there are lots of people out there who know their state (like knowing that guy Bush used to beat up) extremely well and think that what happens there is like it is everywhere. Also, the Martorano measure is most properly used to assess relationships rather than to rank states directly. Probably the reason that no relationships were found in this study was because there is just too much of this random error in these measures.
Finally, as soon-to-be-former editor of SPPQ, let me respond to the last post-- why SPPQ is not free on line. The answer is simple-- it costs plenty to produce and we have to do what we can to get back some of the costs. Pony up the $12 if you want the copy!

[ArchPundit]

Chris,
I do pay my $12 (errr..it may be expired--but I do pay), but also have JSTOR access at work ;)

I generally agree with your point. In this case, I think there may be more to it then simply being random error and that is my point--campaign finance measures and leader influence or control. It strikes me that without that tool, the influence would be far less and autonomy would be greater. Hence it isn't random error so much as omitted variable error. That's, of course, a hypothesis, but one that I think could strengthen the measure if I'm right.

[Lurker]

For Nebraska, just color it in black, or write "NO HOUSE" or something.

[Karl Kurtz]

This message received from Richard Clucas, Portland State University:

Perhaps there is something we learn from weak measures: that they can be improved! The comments by Rich Miller suggest that when evaluating the autonomy of committees one should not only look at the procedures but the politics. Maybe Nancy should have also included a measure of committee "political autonomy" that would assess the extent to which House Speakers and Senate Presidents control committee appointments, committee jurisdictions, bill referral, and similar factors.

As one who has now put together both a formal and informal measure of power, I have often worried that I would get the kind of criticism Nancy has received, so I feel for her. I did appreciate the kind words that the "random political scientist" said about my measure, but I realize it isn't perfect either. Autonomy and power are pretty hard concepts to measure. In the SPPQ article, I relied on the Carey, Niemi, Powell survey, which asked legislators to assess the influence of their leaders. One of the worries I had with the survey is that legislators may have a general perception that their leaders are particularly powerful or particularly weak, which may not be true when looking across states. For example, I have talked to political insiders in Oregon more than once who have proclaimed we have among the most powerful legislative leaders in the nation. When the insiders describe all the powers that the leaders have, I point out that these are common powers in other states, yet they weren't aware of that.
Similarly, New Yorkers routinely believe that they have the most powerful legislative leaders in the nation, but under what basis do they have to assert that claim? The problem I have worried about in relying on a survey in creating my informal measure is that the respondents may be looking at the trees in front of them, without any understanding of the bigger forest. Our goal is to understand the forest. Nancy's measure may be imperfect, but it provides a beginning place to understand the forest. As Miller says later in the blog, perhaps the scores should be adjusted.

Again, thanks for mentioning the exchange. We really do want to have a reality check when we put forward measures, which the responses to your blog have supplied.