Legislative Hearings and Legislative Outcomes
Holly Brasher University of Alabama at Birmingham 1212 University Boulevard, 238 Ullman Building Birmingham, Alabama 35294-3035 email@example.com Presented at the 2005 Annual Meeting of the Southern Political Science Association, January 6-8, 2005 New Orleans, Louisiana
Legislative Hearings and Legislative Outcomes
Abstract: In this study I consider the purposes of legislative hearings. One obvious purpose of hearings is to inform members of Congress about the legislation at hand and resolve problems with the legislation in order to facilitate passage. However, I argue that they are instead intended to communicate with constituents and others outside of the institution. I propose that hearings are costly and that we should expect to observe hearings when the benefit will be the greatest or when the cost is subsidized. I use political actors and configurations of support to create a model of legislative hearings. I argue that hearings and outcomes are motivated by different goals of members of Congress and that hearings do not increase the likelihood that legislation will pass. I use probit analysis to evaluate whether or not a hearing will be held for the legislation and I supplement this with a bivariate probit model using both hearings and the passage of legislation as dependent variables to determine whether hearings and outcomes are independent.
Legislative hearings have an obvious purpose. Testimony about legislation is requested and provided. Members of Congress gather information about a specific piece of legislation and that enables them to make informed decisions about the issue at hand, evaluate the policy solution embodied in the bill, and to move the legislation forward. The institutional purpose of hearings would clearly appear to be to inform member of Congress and to facilitate the legislative process. However, in spite of what appears to be a clear institutional purpose, scholars have proposed that legislative hearings have more than one use. David Truman (1951) suggests that their purpose is threefold. First, he proposes that the institutionalized conflict provides social catharsis. In other words, the function is symbolic in nature. Secondly, in his view hearings serve as a propaganda channel through which public support could be expanded and existing support consolidated or reinforced. This suggests that hearings communicate with constituents and other members of the public. Members of Congress use hearings in an entrepreneurial way to generate support for their preferred outcome. This suggests that a desired outcome exists a priori and the hearing functions as way to ensure that the desired outcome is obtained. And third, he identified what I have noted as the most straightforward purpose, that hearings serve to transmit information to the committee. He adds that this information can be both technical and political (1951, 343-46). The purposes of hearings identified by Truman involve communication that is externally directed toward constituents, interest organizations, and the public at large and communication that is internally directed where the intended recipients are the members of the committee holding the hearing and their congressional colleagues.
Huitt (1954) similarly identifies the dual purposes of hearings. According to Huitt, hearings are “public forums” where “a larger audience participates in the decision,” but they are also “fact finding agencies,” where members of Congress acquire information (342).1 Other studies implicitly assume one purpose or the other, emphasizing exclusively either the publicly (externally) directed communication of legislative hearings or the (internal) provision of information to members of Congress. In studies that assume that legislative hearings are for the benefit of those outside of the institution, hearings are described as window dressing (Berry 1989, 142), and opportunities to claim credit (Davidson and Olezsek 1994). The alternative view of hearings as an opportunity for members of Congress to communicate with fellow members is supported by a more extensively developed literature and theoretical framework. It is based on theories of information asymmetry and assumes that a subset of members of Congress (the committee) may have information about the policy consequences of legislation beyond what is known by their fellow legislators. The information is desired by fellow legislators for the greater certainty it provides about the consequences of policy choices. They acquire this information through specialization in committee (Austen-Smith and Riker 1987; Hardin 1987). Gilligan and Krehbiel (1990) argue that committee members are willing to undertake the costs of specialization and acquire information because their specialization allows them to influence the choices made by less informed non-committee members. Diermeier and
He accepts the premise that information is intended for consumption by members of Congress, however he rejects the characterization of members of Congress as neutral recipients and arbiters and views them instead as an extension of the interests represented by the testimony. In Huitt’s view, little genuine informing or persuasion occurs because members already hold strong opinions.
Feddersen (2000) argue that congressional hearings provide a mechanism that allows committees to transmit information to the floor (52). Those studies that do not use an information asymmetry framework also argue that hearings are useful to members of Congress for the information they provide (Bradley 1980) and find that hearings can communicate both political and technical information to members (Heitshusen 2000). For the purposes of this study, I make no assumptions about the purpose of communication – whether it is to persuade, inform, or signal constituents or colleagues, or whether it communicates technical or political information – but I do make assumptions about whether hearings are intended to communicate internally or externally. I argue that the primary purpose of hearings is to communicate externally with constituents or other interested members of the public. The recent focus has been almost exclusively on inter-institutional signaling. In this paper I present support for the more traditional view of hearings as electorally-motivated communication. What Motivates Committees to Hold Hearings? A limited amount of scholarly attention has been paid to empirically confirming the external communication purposes of legislative hearings. We know that hearings are poorly attended by other members of Congress, and the information provided could be acquired by members elsewhere without the hearing process. An obvious reason why hearings would be intended for external consumption is that they communicate that Congress is active on an issue. Rather than providing information to fellow members of Congress who have ample opportunity to acquire information privately instead of acquiring it publicly though a hearing, I assume that hearings – given their public nature
– have a public purpose. The communication has as its primary goal the development of electoral support for members of Congress. The question of whether hearings are intended for external communication rather than internal communication addressed by considering two processes. First, I consider those factors that influence the likelihood that a hearing will be held. If hearings are associated with those things that legislators would prefer to have visible, such as constituent-centered legislation then this is evidence that supports my claim that hearings are intended for external communication. The second process I consider is whether or not legislation passes. If the primary purpose of hearings is to provide members with information this should increase the likelihood that legislation will pass. If the provision of information to members of Congress is the purpose, during the hearing process conflicts will be resolved, questions will be answered, and alternatives will be chosen. If hearings are internally directed the same factors that increase the likelihood of a hearing should increase the likelihood that legislation will pass. I therefore apply the same model to both processes and compare the relationship between the elements of the model and the two dependent variables. The second dependent variable – whether or not legislation passes – can also be used to test hypotheses produced by the legislative signaling literature (Diermeier and Feddersen 2001). If hearings are intended to influence other members of Congress in way that is preferred by specialists, a hearing should facilitate the passage of the legislation albeit in a direction that is desired by the specialists. Again, if this is the case, the same influences should shape both the likelihood of a hearing and the likelihood that
a bill will pass. Therefore in this study I consider two events – first, whether a hearing is held, and second, whether the legislation passes or does not pass. Model of Hearings I have chosen the Senate as the context for the study rather than the House of Representatives. Senators are less constrained by their committee assignments in their opportunities to work on legislation and the Senate therefore offers a greater opportunity to observe variation in legislator’s discretionary behavior and variation in the legislative process. The full model I propose here includes several components. Each piece of legislation can be characterized by the actors associated with it and by the configuration of support. Beginning with the actors, I consider interest organization activity. The literature on interest organization influence in Congress has produced contradictory findings. Some studies find that interest organizations are able to influence policy outcomes, while others find that their influence is limited or mitigated by other influences. Some of the most useful information produced by these studies indicates that interest organization influence is simply conditional. At times interest organizations are able to have an influence, under other conditions and for various reasons, their influence is limited. One of these studies reports that the impact of interest organization contributions varies by the visibility of the issue (Neustadtl 1990). This intuitively appealing finding produces the expectation that interest organizations will prefer to avoid hearings when possible. Hearings have the potential to attract unwanted scrutiny by the media, constituents, or other observers of congressional activity. Interests will be able to have more of what they want if the public is unaware of the legislation.
Other studies suggest that the role of interest organizations is to provide information to members of Congress. If interest organizations are part of the provision of information internally to members (Leyden 1995), then we should see an increase in the likelihood of holding a hearing as the number of interest organizations active on an issue increases. A positive association between the level of interest activity and the likelihood of holding a hearing will support the understanding of hearings as internally directed communication. Another important actor, certainly, is the president. Should members wish to communicate externally presidential attention associated with an initiative will subsidize the cost of doing so. The president is the focus of national politics and the visibility associated with presidential action means that bills initiated by the administration arrive with a degree of visibility than can easily be raised further by congressional action in the form of a hearing. Because the cost of communicating externally is subsidized, we should expect presidential initiatives to be more likely to be associated with a hearing. Under conditions of divided government, this should be equally true. The majority party in Congress has the opportunity to stack the hearing with unfriendly witnesses to discredit the president’s position. When hearings are understood as externally directed, this produces an expectation that president’s legislation will be more likely to receive a hearing because members of Congress will capitalize on the high level of attention given to all presidential action. The great majority of legislation that is introduced in Congress does not pass consideration and the president’s legislation is no exception (Krutz, Fleisher, and Bond 1998). Peterson (1990) reports that only one-fourth of the president’s legislation passes
in the form preferred by the president. Therefore I have no prediction for whether presidential action will have a positive or negative influence on the likelihood that legislation will pass. Other actors to consider are the members of Congress who serve on the committees of jurisdiction. Because we know that members of Congress seek membership on committees with jurisdiction over legislation of importance to their constituents, committee membership can function as a measure of constituent interests. Bills introduced by members of the committee with jurisdiction should be more likely to receive a hearing because they represent constituent interests and the most direct opportunity for members to benefit electorally from externally directed communication. A complication arises in testing the influence of committee membership on the passage of legislation. Previous literature suggests sponsorship by a member of the committee of jurisdiction will increase the likelihood that the legislation will pass. The authors of studies of the voting decisions of members of Congress have found that members take voting cues from the chairman and senior members of the committee reporting the legislation because of their expertise (Matthews and Stimson 1975, 89-92). This expertise is supported by legislative staff (Sabatier and Whiteman 1985). This suggests that when a member of the committee of jurisdiction sponsors the legislation, their expertise is an internal signal that will help the legislation pass. At the same time, if action by committee members is on behalf of constituents, the particularistic nature of the legislation will not provide a broad base of support and it will therefore be less likely to pass. The expectation for the influence of committee membership on the passage of legislation then is unclear.
Other actors influence the likelihood that legislation will be considered in a hearing, but they are better conceived of as representing configurations of support rather than individual actors. These actors include members of the majority party and cosponsors. With greater agenda control, members of the majority party should have the resources to produce a hearing for their sponsored legislation. Their membership in the majority party suggests that legislation sponsored by a majority party member should be desirable to membership of the majority party (or at least not strongly opposed) and that this legislation should therefore be more likely to be scheduled for a hearing. If it represents goals consistent with majority party preferences then additional salience and communication with the public about the priority will be welcome. The last actors who are part of the configuration of support are the cosponsors of legislation. Mayhew (1974) argues that the intended audience when members cosponsor is constituents and that cosponsorship is low cost position taking (39). Cosponsorship is a quintessentially externally directed act. However, a contradictory conclusion is offered by Kessler and Krehbiel who argue that cosponsorship is an internally directed signal for fellow legislators (1991). They report that early cosponsors may have ideological positions far from the content of the legislation is therefore not necessarily close to the ideological positions their constituents. They derive from this that the purpose of cosponsorship is to signal to fellow legislators rather than to communicate externally to constituents (536). We have an expectation from Mayhew that the number of cosponsors should increase the likelihood of a hearing because legislation with a number of consponsors is legislation that members would prefer to have visible to constituents. We have an expectation from Kessler and Krehbiel that ideological heterogeneity in
cosponsorship should have an influence on passage of legislation, and not hearings. My theory of external signaling is consistent with both of these expectations. A large number of cosponsors suggest that hearings would be desired by the cosponsors, but should the cosponsors be a diverse group of members, this indicates that the hearing will send a message that will be well received by the constituents of some Senators and poorly received by others. I this case a hearing will not be universally desired by all cosponsors and is less likely to be held. As for passage, Young and Wilson (1997) report that consponsorship does not increase the likelihood that legislation will pass, and yet Krehbiel (1991) reports that minority party cosponsorship does increase the likelihood that a bill will pass (Chapter 5). I propose that ideologically diverse sponsorship will be neutral with regard to hearings but will increase the likelihood that legislation will pass. Again, produces hearings are not the same things that produce successful legislation. Data and Methods Which legislation is scheduled for hearings? The answer to this question informs us about their purpose. The cases in this analysis are the bills introduced in the U.S. Senate during the 104th Congress (1995-1996), chosen using a systematic random sample of all legislation (every fifth case), which produced an N of 439 cases. I have chosen the 104th Congress because the legislation data can be complemented by data on interest organization activity for the same period made publicly available by the Lobbying Disclosure Act of 1996 and made available to the scholarly community by Frank Baumgartner and Beth Leech.2 Data on legislation is from the Library of Congress,
The data used here were originally collected by Frank R. Baumgartner and Beth Leech with the support of National Science Foundation grant number SBR 96-31232 and were distributed through the Center for
Thomas: Legislative Information on the Internet web site. Data on partisanship and committee assignments in the 104th Congress is from the Government Printing Office’s Access web site. I analyze two dependent variables in order to understand the role of hearings in the legislative process. The first, of course, is whether or not legislation has received a hearing, but the second is whether or not the legislation was passed by the Senate. If hearings signal to fellow members of Congress, hearings should make it more likely that legislation passes. Both dependent variables are dichotomous and I propose that the outcome for both is shaped by the same set of exogenous influences but not shaped in the same way.3 I also consider the possibility that two dependent variables are not independent. Specifically, I consider the null hypothesis that hearings increase the likelihood that a bill will pass. I used probit to estimate the model but also bivariate probit to evaluate whether or not legislation that has received a hearing differs systematically from that which has not. Bivariate probit is a case selection model similar to tobit and heckit only unlike tobit and heckit is appropriate for two dichotomous dependent variables rather than a dichotomous dependent variable and an interval level dependent variable. The basic logic of sample selection models is that the outcome variable (in this case whether the legislation passes or does not) is only observed if the selection equation dependent variable takes on a certain value (in this case, a hearing is held). An advantage of the model is that if the disturbances are correlated, the bivariate probit estimates will be more efficient than probits estimated separately (Meng and
American Politics and Public Policy at the University of Washington and/or the Department of Political Science at Penn State University. Neither NSF nor the original collectors of the data bear any responsibility for the analysis reported here. 3 I use a dichotomous dependent variable rather than votes in favor – an interval level variable – because most bills are not passed by a recorded vote. Of the 38 that passed, 19 passed by unanimous consent, 9 by voice vote, and 1 failed by unanimous consent. Only 9 had a recorded vote, an interval level measure.
Schmidt 1985). The analysis also produces a value for that indicates whether or not there is a relationship between the two models. The two groups of independent variables represent political actors and their role in the process, and configurations of support. Relevant statistics for the independent variables are listed in Table 1. The first of the political actor variables measures interest organization activity. The Baumgartner and Leech dataset lists all registered in-house and contract lobbyists along with the legislation they sought to influence and the amount of money spent per year for the Senate, the House and the executive branch. I used only those that listed an interest in Senate legislation. Because lobbyists are often active on more than one piece of legislation and the dollar amounts reported are totals rather than the amount spent lobbying for each measure I use the number of organizations active rather than expenditures to create the measure of interest organization activity. The lobbying disclosure forms ask the registrants to list the specific legislation for which they were active, and using the legislation listed, I created a count of the interest organizations active on each bill included in the sample. The number ranges from 0 (the modal category) to 51 interest organizations active (for the National Defense Authorization Act for FY 1997). The second important actor is the president. In the congressional record, legislation that the administration has asked Congress to introduce is noted by the term “by request.” I constructed a measure indicating that a bill is a presidential initiative by coding legislation that included “by request” as 1, 0 otherwise. The third political actors are members of the committee with jurisdiction over the legislation. Again this is simply a dichotomous measure coded 1 if the sponsor of the
legislation was a member of the committee with legislative jurisdiction over the issue and 0 if the sponsor was not a member of the committee. Partisanship and cosponsorship are used to construct the last three measures. I code each according to the sponsor’s membership in the majority or minority party, with Republicans coded 1 and Democrats coded 0. Republicans introduce the majority of the legislation in the sample, specifically 62%. Cosponsors are simply the number of cosponsors associated with each bill. Most legislation is cosponsored, but 35% is sponsored by only one member. Another large percentage, 18%, is sponsored by two members. A final variable is the partisan sponsorship ratio. I use both partisanship and cosponsorship to create this measure of bipartisan support. The ratio is a folded measure and it is the ratio of one party’s members cosponsoring the legislation to the other party’s members. It takes on a value of 0 when cosponsors are all Republicans or all Democrats, .5 when one party has twice as many cosponsors as the other and 1 when the number from each party is the same – and of course all of the values in between. The modal category is 0 (again, much legislation has only one sponsor and these cases are coded 0), and 16% of the cases the ratio is .5 or above. There are 23 cases (5%) that have equal numbers of Republicans and Democrats. Results Results are presented in Table 2. The first thing to note is the value of for the bivariate probit results. Because represents the magnitude of the correlation between the errors of the two equations, we know that if is zero then those factors that influence
whether a bill will receive a hearing are independent of the factors that determine whether legislation will pass or not pass. If is positive, the factors that produce a hearing are
correlated with the factors that influence whether or not a bill passes. In other words, if were positive, we could conclude that bills that receive a hearing are, by virtue of receiving the hearing, more likely to pass. The small absolute value and lack of statistical significance indicate that these two events are independent. This noted, the individual coefficients from the two probit equations give us valuable information about how these processes differ. First, the interest organization activity is positive and significant for both. However, as anticipated, it has a stronger influence on whether or not legislation passes than on whether a hearing is held. Also, for the hearing equation, interest activity is barely significant at the .05 level and for the legislation equation it just misses significance at .01. Legislation of importance to interest organizations may receive a hearing, but their presence does not have a large positive influence and it may receive attention in spite of organization activity rather than because of it. As for the president as an actor in the legislative process, importantly, this variable is dropped from the legislation equation because it perfectly predicts failure. In other words, for this sample, in no case did a presidential initiative pass. At the same time, it has an important influence on whether or not legislation will receive a hearing. The marked contrast in the president’s role supports the conclusion that hearings and passage are independent processes. A committee member as a sponsor increases both the likelihood that legislation will pass and that it will receive a hearing. This is consistent both with my hypothesis about the results and with previous research. I assume that membership on a committee is consistent with constituent interests. From that assumption I hypothesized that bills
sponsored by committee members would reflect constituent concerns and would therefore be more likely to receive a hearing because they represent an ideal opportunity to demonstrate to constituents that members are active on issues that concern them. The results in the legislation equation indicating that committee member sponsored legislation is more likely to pass likely to reflect the reputation for committee member expertise and knowledge that other studies of vote choice have found to influence the decisions of other members of Congress. The data does not allow definitive conclusions about why committee member sponsorship has a positive influence on the probability of a hearing or the probability of passage. Because of the multiple possible explanations, it is only suggestive. However, the result in the hearing equation is consistent with the role I propose for constituent concerns and their influence on externally directed communication in hearings. Majority party member sponsorship also performs similarly in both equations. It is consistent with the hypothesis that members want legislation associated with the majority to be publicly visible. It contributes to the collective good that is the positive public perception of majority party performance. And it is more likely to pass for the same reason. However, for the final two factors – the number of cosponsors and the partisan ratio – we see a telling contrast between the results for hearings and legislation, just as we did with presidential action. Cosponsors, simply by their numbers, can increase the likelihood that legislation will receive a hearing, but they add zero increase to the likelihood that legislation will pass. The contrast between the two effects indicates that cosponsorship is as Mayhew claimed, low-cost position taking (1974, 39). If it were
instrumental for the legislative process, rather than symbolic, we would expect to see some increase in the likelihood of passage with additional sponsors. We see none. What we see instead, is an outcome that provides an additional opportunity for communication. One of the largest differences in influence in the hearing and legislation equations is for the partisan ratio of cosponsors. Sponsorship approaching perfect bipartisanship increases the likelihood of passage significantly, and has an influence on the likelihood of a hearing that is not statistically different from zero. The coefficients for ratio and for the cosponsors contradict the conclusion of Kessler and Krehbiel (1996) who frame their conclusions about cosponsorship as an either/or proposition. In their view, the question is whether cosponsorship serves electoral connection purposes or internal signaling purposes. The results here indicate that it must not necessarily be either/or. Ideological diversity here is associated with greater likelihood of passage, as Kessler and Krehbiel’s theory suggests it will be. But an abundance of cosponsors are associated with a hearing and not with passage. Cosponsorship can serve as to communicate with constituents and also to fellow members.
Conclusion I have proposed here to reconsider our assumptions about the purposes of legislative hearings. Potential audiences include the committee itself, the chamber, and constituents. Does the panel become a fact finding agency when it holds a hearing with information directed toward the committee members? Are hearings internal signals to the median member of the chamber? Do they serve to communicate externally to constituents and others? The results I have presented here support the conclusion that the
purpose of hearings is primarily to communicate with those outside of chamber – constituents and constituencies. Those factors that increase the likelihood that a hearing will be held are the lobbying activity of interest organizations, a presidential initiative, a sponsor who is a member of the committee with jurisdiction over the legislation, a sponsor who is a member of the majority party, and the number of cosponsors. I have argued that hearings are costly. We should expect to find hearings where the benefit will be the greatest or where the costs of external communication are somehow subsidized. The results confirm these conclusions. In the case of presidential initiatives, the president’s high profile provides legislation that comes to the chamber with some of the costs of obtaining attention already paid. In the case of committee members as sponsors and members of the majority party as sponsors, these are priorities of the member’s constituents or the party’s constituency and they bring a greater benefit and therefore justify the costs involved with holding the hearing. A large number of cosponsors also indicate that the legislation is something that will reflect well on those associated with it and this also invites a hearing. The Health Care Fraud Prevention Act of 1995 (21 cosponsors) and the Economic Growth and Regulatory Paperwork Reduction Act of 1995 (28 cosponsors) both reflect a Congress that is working hard to ensure that good triumphs over evil. Interest organization activity does not have the expected effect. However, although interest organizations themselves would probably prefer to have as little public scrutiny as possible, it is members of Congress who decide whether or not hearings will be held, and interest organization presence may reflect conflict that requires a hearing to be resolved.
These basic that I offer as evidence that hearings are publicly directed are supplemented by a bivariate probit model. The small absolute value of the correlation of errors and the lack of statistical significance of the coefficient indicates that that hearings and the passage of legislation are independent phenomena. I also included hearings in the probit model for legislation as an independent variable (in results not reported here but available from the author on request), and it has no predictive value for the outcome of legislation. I offer this as additional evidence that hearings function as an opportunity to communicate externally rather than to signal internally to fellow members of Congress. These results suggest future studies. I chose to limit the cases to legislative hearings. Investigative and oversight hearings would seem to be much more likely to be directed toward an externally audience, however Talbert et al. (1995) find an important role for non-legislative hearings in changing committee legislative jurisdiction. This suggests that they have an internal purpose. Evaluating the process across types of hearings would be useful. I also chose to limit the study to the Senate. The House with its more structured and committee centered legislative process might produce very different results. Much work remains to be done, but the results presented here suggest that earlier insights about the electoral connection were valid.
Table 1: Variable Frequencies and Summary Statistics Maximum Minimum Interest Organization Activity Cosponsors Partisan Sponsorship Ratio Presidential Initiative Committee Member Sponsor Majority Party Sponsor 51 67 1 Yes 11 2.51% 190 43.28% 273 62.19 0 0 0 No 428 97.49% 249 56.72% 166 37.81%
Mean .46 4.43 .15
Standard Deviation 3.36 8.22 .29
Table 2: Probit and Bivariate Probit Results for Hearings and Legislation Dependent Variable Dependent Variable Probit Bivariate Probit Hearing Legislation Hearing Legislation Political Actors Interest Organizations Presidential Initiative Committee Member Sponsor Configurations of Support Majority Party Member Sponsor Cosponsors Partisan Sponsorship Ratio Constant .03* (.02) 1.12** (.41) .30* (.17) .50** (.18) .02** (.28) .16 (.28) -1.84*** (1.9) .05* (.02) -.30* (.18) .50** (.20) .00 (.01) .72** (.27) -1.96*** (.20) .03* (.02) 1.12** (.40) .31* (.17) .50** (.18) .02** (.01) .16 (.28) -1.84*** (.19) -.06 (.14) .16 .05* (.02) --.30* (.17) .50** (.20) .00 (.01) .71** (.27) -1.95*** (.20)
LR test of
p > 0.68
* p <.05, ** p <.01, *** p < .001; Standard errors are in parentheses.
References Austen-Smith, David and William H. Riker. 1987. “Asymmetric Information and the Coherence of Legislation.” American Political Science Review 81(3):897-918. Berry, Jeffrey M. 1984. The Interest Group Society. Boston: Little, Brown. Bradley, Robert B. 1980. “Motivations in Legislative Information Use.” Legislative Studies Quarterly 5(3):393-406.” Davidson, Roger H. and Walter J. Oleszek. 1994. Congress and Its Members. Washington, D.C.: CQ Press. Diermeier, Daniel and Timothy J. Feddersen. 2000. “Information and Congressional Hearings.” American Journal of Political Science 44(1):51-65. Gilligan, Thomas W. and Keith Krehbiel. 1990. “Organization of Informative Committees by a Rational Legislature.” American Journal of Political Science 34(2):531-564. Hardin, John W. 1998. “Advocacy versus Certainty: The Dynamics of Committee Jurisdiction Concentration.” Journal of Politics 60(2 ):374-397. Heitshusen. 2000. “Interest Group Lobbying and U.S. House Decentralization: Linking Informational Focus to Committee Hearing Appearances.” Political Research Quarterly 53(1):151-176. Huitt, Ralph K. 1954. “The Congressional Committee: A Case Study.” American Political Science Review 48(2):340-365. Kessler, Daniel and Keith Krehbiel. 1996. “Dynamics of Cosponsorship.” American Political Science Review 90(3):555-566. Krehbiel, Keith. 1991. Information and Legislative Organization. Ann Arbor, MI:
University of Michigan Press. Krutz, Glen S., Richard Fleisher, and Jon R. Bond. 1998. “From Abe Fortas to Zoe Baird: Why Some Presidential Nominations Fail in the Senate.” American Political Science Review 92(4):871-881. Leyden, Kevin M. 1995. “Interest Group Resources and Testimony at Congressional Hearings.” Legislative Studies Quarterly 20(3):431-439. Matthews, Donald and James Stimson. 1975. Yeas and Nays: Normal Decision-Making in the U.S. House of Representatives. New York: Wiley. Mayhew, David R. 1974. Congress: The Electoral Connection. New Haven, CT: Yale University Press. Meng, Chun Lo and Peter Schmidt. 1985. “On the Cost of Partial Observability in the Bivariate Probit Model.” International Economic Review 26:71-85. Neustadtl, Alan. 1990. “Interest-Group PACsmanship: An Analysis of Campaign Contributions, Issue Visibility, and Legislative Impact.” 69:5490564. Peterson, Mark A. Legislating Together: The White House and Capitol Hill from Eisenhower to Reagan. Cambridge: Harvard University Press. Sabatier, Paul and David Whiteman. 1985. “Legislative Decision Making and Substantive Policy Information: Models of Information Flow.” Legislative Studies Quarterly 10(3):395-421. Talbert, Jaffery C., Bryan D. Jones, and Frank R. Baumgartner. 1995. “Nonlegislative Hearings and Policy Change in Congress.” American Journal of Political Science 39(2):383-405. Truman, David B. 1951. The Governmental Process: Political Interests and Public
Opinion. New York: Knopf. Wilson, Rick K. and Cheryl D. Young. 1997. “Cosponsorship in the U.S. Congress.” Legislative Studies Quarterly 22(1):25-43.