“The bill, completed early Friday and expected to come up for a final vote this week, is basically a 2,000-page missive to federal agencies, instructing regulators to address subjects ranging from derivatives trading to document retention. But it is notably short on specifics, giving regulators significant power to determine its impact—and giving partisans on both sides a second chance to influence the outcome. . .”
—Binyamin Appelbaum, The New York Times on Dodd-Frank, 2010
Lobbying is a business with no deadline. Interest groups lobby to influence agenda setting and voting in Congress, and continue to lobby well after Congress passes legislation.Footnote 1 In this endless game, interest groups participate in different stages of policymaking and use multiple tools in their attempts to achieve influence.Footnote 2 They appear in committee hearings and testimonies, hire commercial lobbying firms and former government officials to contact members of Congress and bureaucrats, meet with regulators, and submit comments during rulemaking processes. Different interest groups and different lobbying activities are observed at different stages of policymaking. Despite the prevalence of lobbying throughout multiple stages of policymaking, there is relatively little research that documents the choices and effectiveness of different lobbying methods across different stages of policymaking.
Scholars have spent copious amounts of time to understand the effects of lobbying on policy outcomes. This endeavor has been challenging because lobbying is inherently a strategic decision. This is particularly true when scholars attempt to estimate the effect of lobbying on votes by members of Congress. First, unlike campaign contributions, it is difficult to measure lobbying contacts at the congressional member level because the Lobbying Disclosure Act of 1995 (LDA), which regulates the domestic lobbying process, only requires lobbyists to disclose the identity of the chamber of Congress or federal agency contacted, not the name of the lawmaker or the bill or topic they were lobbying on. Second, it is well-known that interest groups tend to lobby their allies in Congress, who may already share similar preferences with the interest groups; therefore it is challenging to disentangle the effect of lobbying from inherent, already existing preferences.Footnote 3
Furthermore, the bulk of the literature on the effects of lobbying is focused on lobbying Congress; fewer studies have concentrated on lobbying during the subsequent rulemaking stage, despite rulemaking's importance in implementing legislation. The previous studies that do focus on the rulemaking stage have identified the kinds of participants who are most and least active in comments submissions, and correlate participation with rule changes.Footnote 4 Other questions in the lobbying literature, such as the impact of the “revolving door,” have received attention by scholars in the congressional context,Footnote 5 but have not yet been investigated for outcomes in the rulemaking stage.
This paper advances the literature by studying the lobbying activities of interest groups across two stages of regulatory policymaking—the congressional bill stage and the federal agency rulemaking stage—and by examining whose voices get acknowledged at the end of the process. We track the combinations of lobbying activities used and identify which groups are acknowledged in the final rule. Notably, we consider the “revolving door” in the rulemaking process and investigate whether organizations with former SEC employees are more likely to be acknowledged. Our advantage is a newly collected dataset, explained in detail below, that allows us to directly track references made in an agency rule to specific interest groups who had lobbied on that rule.
The Dodd-Frank Wall Street Reform and Consumer Protection Act (Dodd-Frank, hereafter), passed in July 2010 with around 330 provisions for rulemaking, presents a fitting case study to analyze lobbying in federal agency rulemaking.Footnote 6 We specifically focus on rulemaking activities by the Security and Exchange Commission (SEC), who was responsible for making rules for up to ninety-seven provisions to implement Dodd-Frank. Numerous groups, from corporations to local governments, were involved in lobbying at various stages of the policymaking process.
First, we provide a comprehensive description of the lobbying activities of interest groups across both the congressional bill stage and the rulemaking stage by analyzing which organizations are involved and what types of lobbying strategies—e.g., submitting lobbying reports for lobbying contacts in Congress or the federal bureaucracy, submitting comments to agencies, meeting with federal regulators, hiring former SEC regulators—each organization used to lobby on Dodd-Frank from 2009–14. This enables us to understand how different organizations allocate their resources in different stages of policymaking in their attempt to influence regulatory policymaking.
Next, following studies that have explored interest groups’ activities in the rulemaking process,Footnote 7 we examine how lobbying in the rulemaking stage influences final rules. Previous studies observe which organizations submitted comments and whether the final rule had changes, but do not isolate which specific organizations’ opinions, if any, were taken into account.
We take a new approach that explicitly connects parts of the final rules to specific interest groups by using citations in the final rules—our newly collected dataset—which reference comments and meetings between organizations and the SEC. In the final version of their rules, the SEC includes citations that name the opinions of a selected group of organizations and the medium in which those opinions were transmitted. This allows us to directly connect any considerations the SEC took into account in between the proposed and final version of rules to the specific comments or meetings of specific interest groups. While a citation of an organization in the final rule does not necessarily suggest that an organization's exact preferences are reflected in the final rule, it does indicate acknowledgement of the organization's opinion by federal regulators in the rulemaking process. Thus, our study and our newly collected dataset can identify whether or not the SEC took into account an organization's submitted comment or meeting, allowing us to advance the literature by drawing more specific conclusions on the effectiveness of organizations’ lobbying efforts.
For Dodd-Frank, we find that 2,961 organizations participated in the lobbying process either during the congressional bill stage, the agency rulemaking stage, or both. Corporations and trade associations were the most active, but local governments and members of Congress were also actively involved. While corporations and trade associations lobbied throughout both stages, we find that local governments, not-for-profit organizations, and even individual members of Congress mostly participated in the rulemaking stage exclusively.
We find that the number of lobbying report submissions and the number of meetings with the SEC are strongly associated with the citation of the organization in the SEC's final rule. When we compare comment submissions across various types of organizations, comments submitted by trade associations and members of Congress tend to be more cited in the final rule. When we include a rule fixed effect and organization-type fixed effect, the number of comments submitted by an organization and the frequency of citation of the organization in the final rule are still strongly correlated.
We also investigate whether the “revolving door” influenced the effectiveness of lobbying on Dodd-Frank rulemaking. Research has suggested that lobbyists who were previously employed as House or Senate personal staff, or committee staff, generate higher lobbying revenues over lobbyists who were not.Footnote 8 However, scholars have yet to determine the extent to which a revolving door impacts outcomes in the lobbying of federal agencies. Out of the 2,961 organizations in our data who lobbied the SEC on a Dodd-Frank rule in one way or another, eighty-eight of those groups had employed at least one former SEC regulator. We find that organizations who submitted comments and have at least one former SEC employee are more likely to be cited in the final rules determined by the SEC, and that these results are robust to corporation resources. Combined, our results suggest that the SEC gives higher weight in recognizing opinions from organizations who have issue expertise, capacity to control the agency, or a personal connection to the agency.
Our findings thus provide insight on the lobbying process—who lobbies, how they lobby, and who succeeds in getting acknowledged—across different stages of regulatory policymaking. McCubbins, Noll, and Weingast (Reference McCubbins, Noll and Weingast1987), in their study of bureaucratic agency procedures as instruments of elected officials’ political control, predict that “the politics of the bureaucracy will mirror the politics surrounding Congress and the president.” While we find that there is a wider variety in the types of organizations who lobby the bureaucracy than those who lobby Congress, the types of organizations who are ultimately more likely to succeed in getting their interests acknowledged are similar across the bureaucracy and Congress. In other words, we discover that for the question of which outside groups participate, the politics of the bureaucracy do not mirror the politics surrounding Congress, but for the question of which outside groups get acknowledged, the politics of the bureaucracy do mirror the politics of Congress.
Interest groups and agency rulemaking
While the lobbying that takes place before legislation is passed in Congress attracts much attention, less attention is given to lobbying in the subsequent rulemaking stage, despite rulemaking's critical role in implementing legislation. When Congress passes legislation, it routinely calls for U.S. federal agencies to formulate guidelines and regulations for the new policies through rulemaking. During the rulemaking process, agencies are legally required by the Administrative Procedure Act (APA) to provide notice of the proposed rule in the Federal Registrar and must allow for a notice and comment period before promulgating the final rule (APA Section 553).Footnote 9
During this notice and comment period, the public—either organizations or individuals—can submit written comments containing information, data, or opinions to the agency formulating the rule.Footnote 10 While agencies are not legally required to change a rule in between the proposed version and the final version due to the public comments, if the agency does make modifications from the proposed rule to the final rule, agencies, such as the SEC, mention in citations the comment or meeting that relate to the change being made. Even if a modification is not made, the SEC still has discretion to cite any comment that it finds substantively relevant to the proposed rule.Footnote 11 The final regulatory rules that stem from the agency rulemaking process are legally binding and thus are a form of policymaking.
Like any form of policymaking, agency rulemaking is a process ripe for outside influence—what's more, the above-mentioned notice and comment period is designed to invite and welcome outside opinion. Who, then, influences rulemaking, and are there any groups who unduly affect the outcomes of rulemaking? The literature has argued that the dominant interests of the regulated industry are the drivers behind the scenes during agency policymaking. McCubbins, Noll, and Weingast (Reference McCubbins, Noll and Weingast1987, Reference McCubbins, Noll and Weingast1989) and McCubbins and Schwartz (Reference McCubbins and Schwartz1984) argue that this is because Congress empowers dominant interests in bureaucratic decision procedures. More often than not, the dominant interests are business interests, who have been shown to pursue and receive significant advantages in policymaking.Footnote 12 Perhaps unsurprisingly, then, scholars have also documented that corporations are the most active participants in lobbying the federal agencies during the agencies’ rulemaking phases. For example, Golden (Reference Golden1998), in studying ten rules, found that businesses participate the most in submitting comments during the public comment periods of rulemaking; likewise, Yackee and Yackee (Reference Yackee and Webb Yackee2006) analyzed forty rules made by four different agencies, and found that businesses submitted 57 percent of the comments during the public comment periods of those forty rules.
Not only do corporations participate the most, but scholars have also found that corporations disproportionately influence the federal bureaucracy.Footnote 13 When measuring the amount of change a rule underwent from its initial proposed version to its final version, Yackee and Yackee (Reference Yackee and Webb Yackee2006) found that the four agencies they studied consistently amended their final rules to reflect the desires business interests expressed in their submitted comments. Haeder and Yackee (Reference Haeder and Webb Yackee2015) studied lobbying during the Office of Management and Budget's (OMB) review of rules, and showed that lobbying by business groups exclusively is more likely to result in a rule change than lobbying by public interest groups exclusively. Further, they found that a rule change during OMB review is more likely when more groups lobby and when there is more consensus lobbying (with no lobbying from “the other side”).
More broadly, research is mixed on whether general interest group participation during the rulemaking process affects the resulting rules. Studies such as Golden's (Reference Golden1998), Yackee and Yackee's (Reference Yackee and Webb Yackee2006), and Haeder and Yackee's (Reference Haeder and Webb Yackee2015) have found that interest groups’ comments do, in certain cases, relate to changes made in between the proposed and final rules, though other studiesFootnote 14 have found that interest groups’ comments do not or only rarely do.
A possible reason for the mixed results is that empirical studies have quantified the influence of interest groups on rule content in various ways: automated content analysis software to determine the percentage change in the text from draft to final rule,Footnote 15 manual content comparisons by the author,Footnote 16 or by human coders,Footnote 17 or interviews with agency officials.Footnote 18 However, while these approaches may capture the amount of change in the rules, none of the existing approaches directly attribute any consideration or influence in the rules to the specific interest groups who lobbied or submitted comments. Without a clear link connecting content in the rule to an interest group, there is no certainty that the content was influenced as a result of a certain interest group's submitted comments or other lobbying efforts.Footnote 19
Our approach and data collection allow us to improve upon the literature by filling this gap and directly attributing considerations taken into account by the SEC in between the proposed and final rule to specific interest groups who lobbied during the rulemaking process. By analyzing final rule citations, we can identify who the SEC acknowledges in the formation of the final rule. This allows us to connect interest groups’ lobbying activities to whether their opinions were publicly referenced during the rulemaking stage. While not a perfect measure of influence—a citation does not guarantee that the final rule was actually changed to what an organization lobbied to support—citations nevertheless capture acknowledgment by the agency. As the literature currently does not have a way to systematically and directly attribute interest groups’ lobbying activities to their actual impact on rulemaking, our approach, using citations, is a first step to tackling this gap.
Further, the submission of public comments is not the only avenue that is available to interest groups who wish to lobby agencies during the rulemaking process. Traditional lobbying activities, such as meetings and telephone calls, can still be used to lobby interests, especially if interest groups do not wish to go on the written record with comment submissions. Merely examining comments ignores the other ways that interest groups may lobby agencies; thus, by collecting SEC meeting records with interest groups (meetings both in-person and telephone calls) and attributions to meetings that are present in the SEC final rule citations, we are able to augment our dataset to cover various lobbying activities. Used along with our data on comments, we are able to paint a more complete picture of interest group participation in the rulemaking process and the extent of the SEC's acknowledgement of their opinions in the resulting regulatory rules.
Data and stylized facts
We focus on the SEC's rulemaking for Dodd-Frank. The SEC is one of the most important agencies that oversees the financial market and played an important role in the implementation of Dodd-Frank. When Congress passed the 848 pages of legislation for Dodd-Frank, it had around 330 provisions for rulemaking that it left to the federal agencies. To date, the SEC alone has been responsible for adopting final rules for sixty-seven mandatory rulemaking provisions of Dodd-Frank, completing seventy-two final rules in direct response to mandatory rulemaking provisions and around one hundred rules in total related to Dodd-Frank, thirty-seven of which were finalized by December 2014, the end of our data sample.Footnote 20 For each of the rules, we collected data on how many comments on the rule each organization submitted, how many meetings each organization had with SEC officials regarding that rule, and how many times each organization was cited in the final rule.Footnote 21
From Dodd-Frank's passage in Congress to the end of 2014, 2,337 organizations submitted a total of 4,405 comments on the thirty-seven rules promulgated by the SEC during that time. These comments were provided by the SEC and are all included in our data set. Rules had, on average, ninety-two comments per rule, but the number of comments varies significantly by rule. For instance, rule file S7-7-11 on credit rating reference removal received no comments, but rule file S7-45-10 on the registration of municipal advisors attracted 855 comments. Table A1 in appendix A provides the detailed description about rules and the total comments submitted to each rule.
To provide a complete picture on interest groups’ activity on Dodd-Frank, we collected data on the lobbying activities of various interest groups, both in the congressional bill-passage stage and the agency rulemaking stage, from 2009 to 2015. During this time period, there were 2,961 unique organizations who lobbied either during the congressional stage (by submitting required lobbying reports after having lobbying contact with officials), during the agency rulemaking stage (by submitting comments to the SEC or having meetings with the SEC), or during both. Among these organizations, 22.5 percent only submitted lobbying reports during the congressional stage but did not submit comments or have meetings with the SEC during the rulemaking stage. 68.2 percent did not submit lobbying reports from the congressional stage but did submit at least one comment or had a meeting with the SEC during the rulemaking stage. Further, 9.3 percent were observed submitting both lobbying reports during the congressional stage and submitted comments or had meetings with the SEC during the rulemaking stage.
We categorize organizations into seven different types: corporation, trade association, local government, members of Congress, law or lobbying firm, nonprofit, and individuals.Footnote 22 Table 1 presents the number of lobbying activities categorized by type of organization during the congressional state (number of lobbying reports) and the rulemaking stage (number of comments, meetings with the SEC, and lobbying reports), as well as the number of final rule citations.Footnote 23
Table 2 presents the total number of unique organizations in each category by activity. In all activities, corporations and trade associations are the most active participants in the process. Members of Congress submitted 261 comments in total on the thirty-seven rules; their opinions in these comments were cited 1,140 times. Local governments and law or lobbying firms are more active in comment submission than lobbying report submission, and non-profit groups tend to meet frequently with the SEC. In Table A2 in appendix A, we list the top ten most active groups in each category in each activity.
Citations in the final rules reference organizations and specific comments or meetings. For example, in the final rule on the “Registration of Municipal Advisors” (File No. S7-45-10), the SEC writes,
Many commenters recommended that the municipal advisor registration rules include an exclusion for broker-dealers that is similar in scope to the broker-dealer exclusion under Section 202(a)(11)(C) of the Investment Advisers Act.636 Specifically, these commenters stated that the Commission should exclude from registration broker-dealers that provide advice that is solely incidental to a transaction.637 These commenters generally noted that broker-dealers are already regulated by the Commission and should not be subject to additional or duplicative regulation.
And, furthering the example, the corresponding citation for footnote 637 states:
See, e.g., Union Bank Letter (stating that advice supplied that is “solely incidental to the conduct of his business as a broker or dealer and who receives no special compensation therefor’’ (Section 202(a)(11) of the Investment Advisers Act) should be excluded from the definition of “advice?”; SIFMA Letter I (stating that “broker-dealers providing advice that is solely incidental to a transaction should be excluded from the definition of municipal advisor for the same reason that registered investment advisers are excluded (in some instances): they are already regulated”); Financial Services Institute Letter (stating that broker-dealers should be treated as in the Investment Advisers Act, i.e., where a municipal entity enters into an ordinary brokerage transaction, any incidental advice provided in the scope of that relationship should not require the broker-dealer to register as a municipal advisor).
In this citation, we see that the SEC is citing letters that were submitted as comments during the public comment period from Union Bank, Securities Industry and Financial Markets Association (SIFMA), and the Financial Services Institute.Footnote 24 These three organizations’ efforts are clearly taken into account by the SEC. These footnote citations appear throughout each final rule, and explicitly provide information on what organizations the SEC is acknowledging in the final rules.
We note that a citation does not guarantee that the final rule was actually changed to what an organization lobbied to support. The SEC, in many cases, also cites submitted comments and then explains why it did not follow what those comments called for. Regardless, an organization being cited in the final rule is a signal that its views were heard and taken into consideration—whether or not in an affirmative way—by the SEC in a serious manner. Citations, as such, can be taken as a sign that an organization's lobbying voice was heard.
Types of lobbying and policy change
In this section, we investigate whether different lobbying activities are associated with the number of citations of each organization in the final rule promulgated by the SEC regarding Dodd-Frank. Specifically, we estimate the following model using ordinary least squares:
where i indicates an organization with type j and s indicates the rule. The terms α j and γ s are fixed effects indicating organizational type and the rule, in order to control for unobservable characteristics of organizational types and rules that may affect the number of citations made in the final rule. The variables Comments ijs and Meetings ijs measure the number of comments and meetings that an organization i (with type j) submitted to or had with the SEC regarding a rule s. Reports ij measures the total number of lobbying reports on Dodd-Frank.Footnote 25 We estimate the model using OLS regression.Footnote 26
Table 3 presents the main results. Coefficients for each variable of interest are presented with t-statistics in parentheses. Column (1) includes interaction terms between the number of comments and an indicator for each organization type. Federal agencies are required only to respond to what the courts have characterized as “significant comments” receive during the period of public comments (Carey Reference Carey2013). Given that our outcome measure is the number of citations of comments by each group in the final rule, we include interaction terms between the comments and the group type to examine whether comments from different groups are weighted differently. We also include a rule fixed effect to control the differences across the rules in terms of the attention it draws from various groups and the degree of influence. Column (2) includes both organization type fixed effects and rule fixed effects.Footnote 27 Organization type fixed effects are included to control differences in resources and expertise on issues across different organization types (e.g., corporations vs. nonprofit).
Note: The unit of observation is organization–rule. t statistics in parentheses.
*p < 0.10, **p < 0.05, ***p < 0.01. Standard errors are clustered at the organization level.
In both specifications, the number of meetings with the SEC and the number of submitted lobbying reports are positively associated with the citation of the organization in the SEC's final rule. Comments submitted specifically by corporations, trade associations, members of Congress, and law or lobbying firms tend to be cited more in the final rule when we compare the effect across organization type, as seen in Column (1). For instance, when a member of Congress submitted a comment, that comment was five times more likely to be cited by the SEC in the final rule compared to a comment submitted by an individual/other, which was an omitted, baseline category of the organization. When comparing within each organization type, in Column (2), the number of comments and lobbying reports submitted by an organization, as well as the number of meetings with SEC officers, are strongly correlated with the frequency of citation of the organization in the final rule.
We turn next to the “revolving door” and examine how hiring ex-SEC officials changed an organization's lobbying strategy and the SEC's citation of the organization in the final rule. There is a rich literature on how future career concerns in the private sector could affect regulators’ behaviors.Footnote 28 However, it is not well documented how ex-regulators influence agency policy outcomes when they act as lobbyists. We directly tackle this issue in the following analysis.
The SEC requires that its employees file post-government employment statements if they represent a client before the SEC within two years of leaving the SEC. The Project on Government Oversight (POGO) compiled the statements filed from 2000–2010 by former SEC employees through the Freedom of Information Act.Footnote 29 From this database, we identify all organizations who hired SEC employees who had left the SEC between 2000 and 2010. One caveat is that the SEC only requires this post-government employment statement if the former employee will appear before the SEC in their new position, so the data does not include organizations who hired former SEC employees for positions in which the employee would not go before the SEC.Footnote 30 However, what we can identify are organizations with former SEC employees who interact with the SEC—precisely the “revolving door” phenomenon that we want to investigate.
Eighty-eight organizations who hired at least one former SEC regulator lobbied on Dodd-Frank in either the congressional stage, the rulemaking stage, or both. We examine whether organizations who hired former SEC employees are more likely to be cited in a final rule when they submit comments to the SEC during the rulemaking process. We constructed the variable Hired SEC Employee to indicate whether an organization hired a former SEC employee during 2000–2010 as reported on post-government employment statements. Column (3) in table 3 presents the results of the regression with this indicator variable. Other things equal, organizations with former SEC employees are more likely to be cited in final rules.Footnote 31 Given that the average number of citations for an organization on a rule is 5.1 citations, the magnitude of the coefficient on the variable Hired SEC Employee suggests that overall, having an ex-SEC employee is roughly associated with an increase of over 100 percent in the number of citations of an organization.
Citations in the final rules are a signal that the SEC heard the opinions of the cited organization enough to include an explicit reference, so what can we say from our findings? First, organizations who lobby across both the congressional and rulemaking stages are more successful in getting heard by the regulator—possibly due to the amount of resources these organizations can devote to lobbying on the issue, which in turn may be a reflection of the organization's size or clout and the relevance of the issue to the organization. Second, comments from members of Congress receive many citations in the final rules, suggesting that regulators, such as the SEC, give close attention to the formal written records of those who had legislative control of the policy. Recent studies document that members of Congress constantly engage in direct communications with federal agencies and departments,Footnote 32 in addition to their oversights through their committee activities. Given that any legislator can reprimand agencies by introducing legislation that is unfavorable to the agency or by attacking the agency in the press, agencies like the SEC have a strong incentive to be responsive to members of Congress if they express their opinion through a direct comment submission.
Third, the fact that trade associations and organizations who hired former SEC employees are more likely to be cited suggests that regulators tend to listen more to those who have expertise on issues and/or former connections.Footnote 33 Another potential mechanism that could explain differential citation rates is the legal capacity of the organizations. Litigation and other legal battles have also been argued as tools that interest groups use to influence policy.Footnote 34 It is possible that the SEC pays more attention to comments submitted by organizations who are more likely to engage in legal action with the agency, which are often firms and trade associations. More broadly, while we observe that there is a wide variety of organizations present in lobbying regulatory policy, we find that these specific sets of organizations are more likely to get their interests acknowledged by the bureaucracy—a first step in influence.
These findings have important implications in the inequality of influence that different groups have in policymaking. Although some scholars argue that allowing citizens to submit rulemaking comments through online submissions (electronic commenting) achieves more “regulatory democracy” by increasing public participation in the rulemaking process,Footnote 35 our results imply that the presence of individual citizens and nonprofit organizations in the rulemaking process does not necessarily lead to the same level of recognition from federal agencies that trade associations or corporations enjoy. This inequality in recognition could be driven by the difference in comment quality, given that individuals’ comments generally only express support or opposition for a proposed rule rather than providing detailed suggestions or information,Footnote 36 or could be driven by the difference in personal connections with federal bureaucrats, since corporations and trade associations more often hire ex-regulators. To fully understand the source of unequal recognition of different groups by federal agencies in the rulemaking process, future research needs to both carefully and systematically examine the content of comments and final rules and how bureaucrats respond content-wise to comments in the final rule.
To view supplementary material for this article, please visit https://doi.org/10.1017/bap.2018.27