The Making of Financial Regulation - Voting on. the U.S. Congress

Transcription

1 The Making of Financial Regulation - Voting on the U.S. Congress João Rafael Cunha University of Cambridge & Graduate Institute, Geneva Abstract This paper studies the voting patterns of congressmen on financial regulation between 1991 and It uses the most comprehensive dataset ever assembled on campaign contributions from the financial sector and it is the first study taking a long-term perspective and that controls for the business cycle. It shows that campaign contributions are the strongest driver of congressional voting. Keywords: Financial Regulation; Political Economy JEL: G2, G28, D72, P16 I would like to thank the helpful discussions and thoughtful comments from Ugo Panizza, Rui Esteves, Klaas Mulier, Tony Moore, Rahul Mukherjee, Richard Varghese, Yuan Zi and the participants in the Swiss Society of Economics and Statistics Annual Congress, ESH Seminar in Oxford University, The Political Economy of Regulation Workshop at the European University Institute, Belgian Financial Research Forum. All renaming mistakes are mine. The research leading to these results has received funding from the People Programme (Marie Curie Actions) of the European Union s Seventh Framework Programme FP7/ / under REA grant agreement n Author s contact: jrsd2@cam.ac.uk 1

2 Introduction There is a vast literature on financial regulation dedicated to the determination of optimal financial regulation. In a recent seminal work, Brunnermeier et al. (2009) present the basic guidelines for financial regulation based on the experience of the financial crisis of Additionally, Barth, Caprio, and Levine (2008) evaluate and suggest different types of bank regulation based on a worldwide analysis. However, misguided regulation is frequently considered one of the main causes of costly financial crises. In the recent crisis of , Eichengreen (2008) criticized regulators for not adapting banking regulation after the elimination of the Glass- Steagall Act. Studying another event, Romer and Weingast (1991) argue that the severity of the Savings & Loans (S&L) crisis could be largely attributed to the policy of regulatory forbearance established by Congress between 1985 to 1987 that allowed for massive gambling for resurrection and ended up just generating even more losses. Thus, there appears to be a general academic consensus regarding the bedrock principles of sound financial regulation. At the same time, there are numerous examples of financial crises plausibly caused by poor regulation. To understand this phenomena, it is of particular interest to investigate the determinants of financial regulation, what I label as the production of financial regulation. This paper tries to bridge the gap between the two aforementioned strands of literature by studying the production of financial regulation and analyzing how this varies between bills that loosen or tighten regulation. For this purpose, I look at the role different inputs play in this production and try to find out which are the relevant variables. The inputs under consideration here are campaign contributions, 2

3 constituency interests, business cycle and ideology. For this study, I compiled a unique and comprehensive dataset with variables capturing all these factors. To measure corporate interests, I collect data on campaign contributions from financial companies. I cover the longest possible period for which this data is available. Despite the fact that lobbying contributions only became obligatory to report in 1995, I collect it since This variable is important to include because the literature on political economy finds that corporate or business PACs tend to donate money based on a pragmatic exchange model, in which interest groups contribute to candidates who are likely to return favors. This contrasts with the model used by labor or ideological groups, who usually support like-minded candidates (for a discussion on the topic see Potters and Sloof (1996)). When studying the role of interest groups on voting by lawmakers, a concern may arise that the effect of campaign contributions is merely capturing the interests of their constituencies through the employment channel. For this purpose, I include the share of population in the congressional district (if the legislator is in the house of representatives) or state (if the legislator is a senator) employed in the financial industry. Additionally, I also collect data on GDP to control for the economic cycle. Then, I add roll call voting records on all the bills that tried to changed financial regulation since Finally, following the political science literature, I use the DW-Nominate score to measure ideology. Ideology is used as a control of an alternative explanation for legislators behavior (Poole and Rosenthal (1996)). Dawning upon the rich dataset that I assembled, I examine whether the voting 3

4 patterns of congressmen can be explained by the aforementioned factors. I also explore the heterogeneity between bills that attempted to loosen or tighten regulation. I find that campaign contributions are the strongest driver of congressional voting for financial deregulation. They increase in the likelihood of voting in favor of deregulation. This paper has four contributions to the literature on the production of financial regulation. This is the study that covers the longest period of time and largest number of bills on financial regulation for which we have data on campaign contributions. This allows me to generalize on what are the factors that most influence politicians when they vote on financial regulation. Additionally, it enables me to look at different points in the business cycle. So far, the literature on the political influence of financial companies on their own regulation has only looked at individual bills, a small subset of bills or bills voted on a small window of time (See, for instance, Mian, Sufi, and Trebbi (2010); Igan and Mishra (2011); Mian, Sufi, and Trebbi (2013)). This is also the most comprehensive dataset assembled on campaign contributions from the financial sector. This allows me to also generalize on the influence this variable has on the production of financial regulation. I believe this is important to better understand the purpose and influence of campaign contributions on financial regulation, because there is evidence that this channel influences legislative voting (Mian, Sufi, and Trebbi (2010), Igan and Mishra (2011) and Mian, Sufi, and Trebbi (2013)), supervisory enforcement (Lambert (2014)) and risk taking (Igan, Mishra, and Tressel (2011)). Furthermore, this is the first attempt to empirically capture the influence of cycles 4

5 in financial regulation. It follows from the work of Rajan (2009) and Coffee Jr (2011), who point out that there are fluctuations in the oversight of financial regulators. They suggest that regulation and oversight are not constant, but rather follow the passion of the public for financial reform. This means that it increases after an episode of financial turmoil. But then, as the financial markets and the economy start to recover, there is frequently a momentum for deregulation. Moreover, to the best of my knowledge, this is the only study that compares the influence of all four aforementioned factors (private interests, constituency interests, ideology and cycles) in any field of regulation, not only financial. Finally, this paper contributes to the debate on the motives of financial regulation. Some suggest that regulation exists due to public interest, because it corrects market imperfections, such as incomplete information and monopolies (see Kaufman and Kroszner (1997)). Others claim that it is driven by private interests (as suggested by Stigler (1971) and Krueger (1974)). Lastly, others propose it exists due to ideological reasons (Poole and Rosenthal (1996)). For contributions on this debate see Vidal, Draca, and Fons-Rosen (2012); Braun and Raddatz (2009); Kroszner et al. (1999); Calomiris and Haber (2014); White (1982); Calomiris and White (1994); Kroszner and Strahan (1999); Ramirez and De Long (2001). The next section presents some background for the empirical specification, which is followed by a description of the major pieces of legislation legislation used in this study. The next section provides a description of the data used and the summary statistics. Then, I present a brief explanation of the empirical strategy used and the results obtained. The last section concludes this paper. 5

6 Background Banking crises are very costly 1 and often considered to be the cause of misguided regulation, as argued by Eichengreen (2008), who criticizes the fact that banking regulation was not adapted after the elimination of the Glass-Steagall Act. In the aftermath of the recent financial crisis, many sectors of the U.S. population asked for a regulatory reform of the financial sector, particularly the legislation guiding the activities of banks. As a response, a few reforms were implemented. However, the criticisms continue, mostly because they claim the reforms are insufficient and the legislators are accused of caving to the political influence of the financial industry. Such claim was mostly conveyed by the fourth estate, relying usually only on anecdotal evidence. According to a report in the Wall Street Journal by Simpson (2007), the Ameriquest Mortgage and the Countrywide Financial, two of the largest mortgage lenders in the United States, spent $20.5 and $8.7 millions of dollars respectively in political donations, campaign contributions, and lobbying activities from 2002 to 2006, to prevent the enactment of anti-predatory lending legislation and other similar bills. Those claims were not only made by journalists. Johnson (2009) also criticises the powerful financiers elite and its political influence, exerted both through campaign contributions and through the Wall Street - Washington corridor. Such revolving door between Wall Street and congress was severely lambasted during the financial crisis, particularly the political power obtained by several Goldman Sachs alumni, 1 See Hoggarth, Reis, and Saporta (2002); Laeven and Valencia (2008) for estimations of the those costs. 6

7 dubbed as Government Sachs (Creswell and White (2008)). Even after the beginning of the financial crisis, in a time when the financial system was in great distress, the banks were still fighting strongly against tighter regulation. In 2009, The New York Times reported that the nine largest participants in the derivatives market, including JPMorgan Chase, Goldman Sachs, Citigroup and Bank of America, joined forces and established a new lobbying organization, the CDS Dealers Consortium, to influence the legislation on derivatives being discussed in congress (Morgenson and Van Natta Jr (2009)). Furthermore, Democratic Representative Colin C. Peterson, chairman of the House Agriculture Committee, the body that oversees the Commodity Futures Trading Commission, is quoted in the same article expressing a popular frustration regarding the inability to enact tighter regulations on banks: The banks run the place, I will tell you what the problem is they give three times more money than the next biggest group. It s huge the amount of money they put into politics. 2 Legislation For this study, I tried to include every bill voted in Congress that in some way changed financial regulation between 1991 and A full list of the bills that will be studied, as well as additional details, is provided in the appendix. Since the 1980s, the U.S. has seen a vast deregulation of its financial regulatory framework. However, as a consequence of the S&L crisis, that led to the resolution of nearly 1300 S&Ls and 300 banks from 1980 to 1994, Congress passed the 2 This The New York Times article reported that securities and investment companies spend $152 million in political contributions from 2007 to 2008 through their political action committees. 7

8 Government-Sponsored Housing Enterprises Financial Safety and Soundness Act of 1991 and the Federal Deposit Insurance Corporation Improvement Act of These bills attempted to put a greater focus on risk-based capital levels and increase the transparency of depositary institutions. Nonetheless, the S&L crisis did not stop the momentum and popularity of deregulation. This continuing trend led to the Riegle-Neal Interstate Banking and Branching Efficiency Act of 1994 and the National Securities Markets Improvement Act of The former relaxed the interstate restrictions on branch banking, while the latter relaxed the Federal securities laws. The century ended with the the famous and vastly discussed Gramm-Leach-Bliley Act, which formally repealed the Glass- Steagall Act. Another important piece of legislation at this time was the Commodity Futures Modernization Act of It exempted financial derivatives from regulation, meaning that products such as credit default swaps, which were important in the propagation of the financial crisis of , were virtually unregulated. The first two years of the new millennium saw a couple of restrictive bills being passed as a consequence of the Enron scandal. The most emblematic was the Sarbanes-Oxley Act of 2002, which established stricter rules for accounting and auditing practices. After this, the early 2000s saw almost annual financial services regulatory relief acts loosening even further the regulations on bank branching, mergers and reserves. This allowed banks to increase their leverage and grow rapidly, creating several so called Too Big to Fail institutions. 8

9 Even the bills that failed to pass to become law were mainly targeted at loosening financial regulation. Examples of those cases were the Anti-Predatory Lending Act of 2000 and the Mortgage Reform and Anti-Predatory Lending Act of 2007, which targeted predatory lending and meant to facilitate litigation for lending practices considered unfair. Consumer protection was also not strengthened when the Consumer Mortgage Protection Act of 2000 failed to pass. Moreover, since the early 2000s, one also notices a strong government policy push towards a mortgage credit expansion, especially for the underserved population. This meant an expansion of subprime mortgage credit. An early example of this is the American Dream Downpayment Act passed in However, this bill was so popular and enjoyed so much bipartisan support that it passed unanimously and without any roll call vote. Nonetheless, other important bills related with mortgage credit were voted in the U.S. Congress in this period. As examples, I highlight the Federal Housing Finance Reform Act of 2005, the FHA Manufactured Housing Loan Modernization Act of 2006 and the Expanding American Homeownership Act of Such an enormous built up in the subprime credit eventually led to a large amount of defaults and a credit crunch. The aftermath of this crisis led to the enactment of the Emergency Economic Stabilization Act. It gave the U.S. Treasury Department up to $700 billion to recapitalize the banks by direct acquisition of distressed mortgage backed securities or new equity. Another very important piece of legislation created in response to the crisis was the Dodd-Frank Wall Street Reform and Consumer Protection Act. It included 9

10 several provisions, most notably on consumer protection, the end of Too Big to Fail bailouts, the creation of an advance warning system and the increase of transparency of exotic financial products and of credit rating agencies. Despite the regulatory push witnessed in the aftermath of the crisis, the regulatory tide quickly turned leading to an erosion of the bills passed during the crisis. This is particularly true for the Dodd-Frank Act. Examples include the Business Risk Mitigation and Price Stabilization Act of 2012, the Promoting Job Creation and Reducing Small Business Burdens Act and amendments H.R and H.R (113th) from Among other things, these bills loosened the margin requirements for over-the-counter swap trades, allowed bank holding companies with participation on insurance companies to avoid minimum capital requirements under the Federal Reserve rules and prohibited the application of the Volcker rule before July 21, Data Lobbying contributions only became obligatory to report in 1995 with the Lobbying Disclosure Act of Despite it, I was able to collect data on campaign contributions since Hence, the data for this study covers the 102nd Congress through the 113th Congress spanning from 1991 to For this period, I went over the details of each bill and classified them into two types, those promoting more restrictive regulation in the financial sector and those deregulating this same sector. This categorization was made based on the contents of each bill. This information can be 10

11 found in the Library of Congress website ( This study includes both bills that passed and failed in congress. The classification of bills is relatively unambiguous. It is labelled as a restrictive bill if it limits the activities of financial companies, while looser bills promote deregulation in the financial sector. This way, one expects that financial companies would oppose the former and favour the later. The roll call records for all the votes from all senators and representatives can be found at a detailed website maintained by Keith Poole. Furthermore, in this website, we can also find some data on political variables, such as party affiliation. I should add that none of the bills included in this paper was unanimously approved or rejected. The first set of explanatory variables covers financial interests of the legislators, including campaign contributions. The politically targeted activities of lobbyist are usually done through political action committees, commonly known as PACs. The data on their campaign contributions is available on the website of the Center for Responsive Politics ( There, I collected the data on donations from companies classified by the Center for Responsive Politics on the Finance, Insurance and Real Estate industry (FIRE). The data on contributions has to be collected on the campaign cycle before the Congressional session. Hence, this data was collected since 1988, even though the early years have a large number of missing observations. Then, I collect data on the constituency interests to measure the influence of the constituency. My measure of choice in this case is the percentage of people working 11

12 in the financial industry in each congressional district (in the case of representatives) or state (for senators). This data was collected from the the United States Census Bureau at This variable is, unfortunately only available for 2000 and from 2006 onward. To control for cyclicality, I collected state level GDP growth rate from the Bureau of Economic Analysis. Political ideology is not the most straightforward variable to measure. Nonetheless, given its importance in this study, it should be included. In this analysis, I will use the most standard measure in the political science literature, which is the DW- Nominate ideology score developed by Poole and Rosenthal (1985). Both measures of central tendency reported for the DW-NOMINATE scores (ideology) are positive, which means that the Congresses under study seem to be ideologically more economically conservative. In this measure, a score closer to 1 can be described as a conservative, whereas a score closer to -1 is usually described as a liberal. Finally, I use political control variables, which are for the most part straightforward in measurement and collection (take party affiliation for instance) and are available in the aforementioned political database The data on committee assignments was collected from data_page.html, a website maintained by political science professor Charles Steward III. The construction of the dataset for this paper was a challenge. Matching the data collected from the different sources was an onerous process due to the very different ways in which political candidates are identified by each source. Hence, I had to be 12

13 very careful and meticulous in this process. Table 1 presents summary statistics at the congressmen level. The fraction of the population working in the financial industry is not very large. More interestingly, observing this last variable, one also notes the lack of link between constituency interests through employment and campaign contributions from the financial industry. Finally, we have the set of political variables. The variable party takes the value 1 if the congressmen is a democrat and zero if a republican. Hence, one can observe that there were a few more republicans than democrats during the period analyzed in this study. Senior Party refers to members of the parties leadership. This includes speakers, chairmen and whips. For these aforementioned legislators, this variable takes the value 1. Committee Seniority counts the number of terms a congressmen has spend as a member of a committee. Empirical Strategy The methodology that will be used in this study is a binary dependent variable regression with panel data to help explain which are the factors that determine the votes by the legislators. The benchmark regression used is the following linear probability model 3 3 The use of a linear probability model in congressional voting is discussed formally in Heckman and Snyder Jr (1996). 13

14 V ib = α + β 1 CC i + β 2 P opf inance i + β 3 GDP + β 4 DW nom i +γcontrols ib + ε where V is equal to 1 if legislator i voted aye on loosening bill B and zero otherwise; on bills to add regulation V takes the value 0 if legislator i voted aye on bill B and one otherwise; CC i is the log of the campaign contributions to legislator i. The measure of constituent interest is P opf inance i, which is the share of population working in the financial industry in the congressional district, in the case of the House of Representatives, or state, for the Senate, of the lawmaker i. GDP s is the annual growth rate of the legislator s state s. DW nom i measures of political ideology of congressmen i. Finally, the next group of variables are controls aimed at capturing the influence of political variables in the voting behavior of legislators. These variables include party affiliation, number of terms served in congress and seat on the financial committee or not. Results One of the main advantages of this study is the ability to isolate the effects of interest groups, constituency interests, ideology and business cycle. Table 2 presents the regression estimation results regarding voting for deregulation. In this case, 1 is a vote in favour of bills that loosen financial regulation or against bills that increase 14

15 regulation. Campaign contributions appears with a positive sign, which indicates that an increase in campaign contributions is associated with an increase in the likelihood of voting for deregulation. Column 1 displays the results for a linear probability model. In this case, the coefficient estimate means that a one standard deviation (1.16) increase in the log of contributions in the previous cycle causes a 4.4 percentage points increase in the probability of voting in favor of deregulation. The results using a probit or logit model are very similar. GDP growth is associated with an increase in the likelihood of voting for deregulation. A one percentage point increase in GDP growth causes a 7 percentage points increase in the probability of voting in favor of deregulation. This means that increases in the growth of the economy lead to more votes for a less regulated financial sector. Ideology also causes an increase in the likelihood of voting for a less regulated financial sector. Constituency interests, however, are not associated with voting for financial regulation. An advantage of this study is the different votes on the same legislation. This analysis reveals the most important factors pushing marginal legislators to change their initial vote. Table 3 presents the results for the same bill voted in different congresses. This means that there is an election, and hence, a campaign contribution cycle between the two votes. This allows us to better isolate the effects of campaign contributions on congressional voting. The main determinant of vote switching are campaign contributions. I believe this is a very strong indicator of the importance 15

16 of this variable in the voting process. Next, I use nearest neighborhood matching to estimate the effect of receiving campaign contributions from the financial industry. I match politicians based on ideology, constituency employment and party affiliation. The results are presented in table 4. These are similar to the ones obtained before. Once again, legislators who receive campaign contributions from the financial industry are 4.6 percentage points more likely to vote for deregulation. This effect is even stronger if we only consider bills that loosen financial regulation. Next, I look at the ideology interaction with campaign contributions, GDP growth and constituency employment. The goal is to analyze if politicians that are ideologically extreme are more or less sensitive to these factors. The results are presented in table 5. Column 1 displays the interaction term between ideology and campaign contributions, which is positive. This indicates that legislators with a more conservative ideology are more responsive to campaign contributions. The interaction term between ideology and GDP growth is negative, which implies that lawmakers who are ideologically more conservative are less responsive to GDP growth. Finally, the interaction term between ideology and constituency employment is negative. This means that legislators with a more conservative ideology are less responsive to constituency employment. Robustness Checks In this section, I am going to perform a multitude of robustness checks to address potential problems with the empirical strategy used. 16

17 A first concern would be that lawmakers are not responding to campaign contributions from finance and that the results obtained for this variable in the previous regressions would also be obtained if instead we had used campaign contributions from other industries To address this, I ran the same regressions, but instead of using campaign contributions from the financial sector, I used the all campaign contributions excluding the ones from the financial sector. This results are presented in table 6. We can observe that contributions from other sectors do not seem to be relevant to the voting in bills concerning financial regulation. In the core specification, I use the log of campaign contributions on the last election cycle as a measure of the special interest of the financial sector. Table 7 presents the results with different measures and ways to calculate campaign contributions for all bills. The average and totals of the last 3 electoral cycles and of all the previous cycles combined attempt to capture the relationship built between the financial industry and the legislator. The results are similar to the ones obtained in the baseline results. This variable is always positive and almost always statically significant. This is further evidence that campaign contributions are a driver for deregulation in the financial system. Conclusion This paper analyzed the production of financial regulation and the different inputs of this process. Studying the voting patterns of congressmen in bills that restrict or 17

18 loosen financial regulation between 1991 and 2014, I find that campaign contributions are the main driver of congressional voting and they lead to an increase in the likelihood of voting in favor of deregulation. This conclusion is in line with a recent strand of literature that shows that special interest pressure played a very important role in the deregulation of the financial system, particularly the mortgage credit market (See Igan, Mishra, and Tressel (2011) and Mian, Sufi, and Trebbi (2013)). Additionally, Mian, Sufi, and Trebbi (2010) also find that constituency interests play a role in voting on financial regulation. Nonetheless, this paper expands on this literature by including a more comprehensive dataset on campaign contributions, covering a larger period of time and bills and controlling for the business cycle. This allows us to generalize on what are the factors relevant for the production of financial regulation. This way, we are able to identify the factors that contribute at all moments for this production. References Barth, James R, Gerard Caprio, and Ross Levine Rethinking bank regulation: Till angels govern. Cambridge University Press. Braun, Matias and Claudio E Raddatz Banking on politics. World Bank Policy Research Working Paper Series, Vol. Brunnermeier, Markus Konrad, Andrew Crockett, Charles AE Goodhart, Avinash 18

19 Persaud, and Hyun Song Shin The fundamental principles of financial regulation, vol. 11. Centre for Economic Policy Research London. Calomiris, Charles W and Stephen H Haber Fragile by Design: The Political Origins of Banking Crises and Scarce Credit. Princeton University Press. Calomiris, Charles W and Eugene N White The origins of federal deposit insurance. In The regulated economy: a historical approach to political economy. University of Chicago Press, Coffee Jr, John C Political Economy of Dodd-Frank: Why Financial Reform Tends to be Frustrated and Systemic Risk Perpetuated. Cornell L. Rev. 97:1019. Creswell, Julie and Ben White The Guys from Government Sachs.. The New York Times 19. Eichengreen, B Origins and Responses to the Crisis. University of California, Berkeley. Heckman, James J and James M Snyder Jr Linear probability models of the demand for attributes with an empirical application to estimating the preferences of legislators. Tech. rep., National bureau of economic research. Hoggarth, Glenn, Ricardo Reis, and Victoria Saporta Costs of banking system instability: Some empirical evidence. Journal of Banking & Finance 26 (5): Igan, Deniz and Prachi Mishra Three s company: Wall Street, Capitol Hill, and K Street.. 19

20 Igan, Deniz, Prachi Mishra, and Thierry Tressel A fistful of dollars: Lobbying and the financial crisis. Tech. rep., National Bureau of Economic Research. Johnson, Simon The quiet coup. The Atlantic 52. URL theatlantic.com/doc/200905/imf-advice. Kaufman, George G and Randall S Kroszner How should financial institutions and markets be structured? Analysis and options for financial system design.. Kroszner, Randall S and Philip E Strahan What drives deregulation? Economics and politics of the relaxation of bank branching restrictions. The Quarterly Journal of Economics 114 (4): Kroszner, Randy et al Is the Financial System Politically Independent?: Perspectives on the Political Economy of Banking and Financial Regulation. Citeseer. Krueger, Anne O The political economy of the rent-seeking society. The American economic review 64 (3): Laeven, Luc and Fabian Valencia Systemic banking crises: a new database. IMF Working Papers :1 78. Lambert, Thomas Lobbying on Regulatory Enforcement Actions: Evidence from Banking. Available at SSRN Mian, Atif, Amir Sufi, and Francesco Trebbi The political economy of the subprime mortgage credit expansion. Tech. rep., National Bureau of Economic Research. 20

21 Mian, Atif Rehman, Amir Sufi, and Francesco Trebbi The Political Economy of the US Mortgage Default Crisis. American Economic Review 100 (5): Morgenson, Gretchen and Don Van Natta Jr In Crisis, Banks Dig In for Fight Against Rules. The New York Times: May 31. URL /06/01/business/01lobby.html?pagewanted=all&_r=0. Poole, Keith T and Howard Rosenthal A spatial model for legislative roll call analysis. American Journal of Political Science : Are legislators ideologues or the agents of constituents? European Economic Review 40 (3): Potters, Jan and Randolph Sloof Interest groups: A survey of empirical models that try to assess their influence. European journal of political economy 12 (3): Rajan, Raghuram G The credit crisis and cycle-proof regulation. Federal Reserve Bank of St. Louis Review 91 (5): Ramirez, Carlos D and J Bradford De Long Understanding America s hesitant steps toward financial capitalism: Politics, the Depression, and the separation of commercial and investment banking. Public Choice 106 (1-2): Romer, Thomas and Barry R Weingast Political foundations of the thrift debacle. In Politics and Economics in the Eighties. University of Chicago Press,

22 Simpson, Glenn R Lender Lobbying Blitz Abetted Mortgage Mess. Wall Street Journal :12 31URL SB html. Stigler, George J The theory of economic regulation. The Bell journal of economics and management science :3 21. Vidal, Jordi Blanes I, Mirko Draca, and Christian Fons-Rosen Revolving door lobbyists. The American Economic Review 102 (7): White, Eugene N The political economy of banking regulation, Journal of Economic History 42 (1):

23 Appendix Tables Table 1 presents the summary statistics. Table 1: Summary Statistics count mean sd min max Party Ideology Member of a Financial Committee Senior Member of a Financial Committee Senior Party Member Real GDP growth Total Campaign Contributions in the Last Election Cycle e+07 Total Campaign Contributions from Finance in the Last Election Cycle Percentage of Campaign Contributions from Finance in the last cycle Fraction Population working in Finance, Insurance and Real Estate

24 Table 2 presents the coefficients estimates relating voting patterns to ideology, campaign contributions, GDP growth and constituency employment. The dependent variable takes the value 1 if the lawmaker voted in favor of a bill proposing looser financial regulation or against a bill proposing tighter financial regulation, and zero otherwise. Political controls include being in a finance committee, committee seniority and party seniority. Column 1 presents the estimates using a linear probability model. Column 2 presents the estimates using a probit model. Column 3 presents the marginal effects of the probit model. Column 4 presents the estimates using a logit model. Column 5 presents the marginal effects of the logit model. Robust standard errors are presented in parentheses. Table 2: Regressions Results - All Bills (1) (2) (3) (4) (5) LPM Probit MFX Probit Logit MFX Logit b/se b/se b/se b/se b/se main Ideology 0.140*** 0.395*** 0.140*** 0.639*** 0.138*** (0.01) (0.03) (0.01) (0.06) (0.01) ln(campaign Contributions *** 0.130*** *** 0.213*** *** from Finance) (0.01) (0.02) (0.01) (0.02) (0.01) Real GDP growth *** 0.200*** *** 0.332*** *** (0.00) (0.01) (0.00) (0.01) (0.00) Fraction Population working in Finance, Insurance and Real Estate (0.28) (0.78) (0.28) (1.28) (0.27) Constant *** *** (0.06) (0.17) (0.28) R-sqr Pseudo R-sq N A */**/*** next to coefficient indicates significance at the 10/5/1% level. 24

25 Table 3 looks at legislators who change their voting on the same bill. It presents the coefficients estimates relating voting patterns to campaign contributions, GDP growth and constituency employment. Political controls include being in a finance committee, committee seniority and party seniority. Column 1 presents the estimates using a linear probability model. Robust standard errors are presented in parentheses. Table 3: Regressions Results - Switching Across Different Congresses (1) Switching b/se ln(campaign Contributions 0.166*** from Finance) (0.04) Real GDP growth (0.02) Fraction Population ** working in Finance, Insurance and Real Estate (1.97) Constant * (0.49) R-sqr N 183 A */**/*** next to coefficient indicates significance at the 10/5/1% level. 25

26 Table 4 presents the nearest neighborhood matching average treatment effect estimates of receiving campaign contributions from the financial industry onto voting on financial regulation. Column 1 presents the results for voting in favor of fewer legislative constraints, which include voting in favor of bills introducing looser legislation and voting against bills proposing tighter legislation. Column 2 presents the results for voting against bills proposing tighter legislation. Column 3 presents the results for voting in favor of bills introducing looser legislation. Table 4: Treatment Effects Estimates based on Nearest Neighborhood Matching (1) (2) (3) Vote Vote Looser Vote Tighter b/se b/se b/se ATE r1vs0.contributions *** 0.102*** (0.01) (0.01) (0.02) N A */**/*** next to coefficient indicates significance at the 10/5/1% level. 26

27 Table 5 presents the coefficients estimates of the interaction terms. Political controls include being in a finance committee, committee seniority and party seniority. Column 1, 4 and 7 present the estimates for all bills using a linear probability model. Column 2, 5 and 8 present the estimates for bills loosening financial regulation using a linear probability model. Column 3, 6 and 9 present the estimates for bills tightening financial regulation using a linear probability model. Robust standard errors are presented in parentheses. Table 5: Regressions Results - Interaction Terms (1) (2) (3) Vote Vote Vote b/se b/se b/se Ideology *** 0.166*** 0.208*** (0.12) (0.01) (0.04) ln(campaign Contributions *** *** *** from Finance) (0.01) (0.01) (0.01) Real GDP growth *** *** *** (0.00) (0.00) (0.00) Fraction Population working in Finance, Insurance and Real Estate (0.28) (0.28) (0.30) (Ideology score) x *** (ln(camapaign contributions)) (0.01) (Ideology score) x (Real *** GDP growth) (0.01) (Ideology score) x * (Constituency employment) (0.56) Constant 0.123** (0.06) (0.06) (0.06) R-sqr N A */**/*** next to coefficient indicates significance at the 10/5/1% level. 27

28 Table 6 presents the coefficients estimates relating voting patterns to campaign contributions made by non-financial firms. Political controls include being in a finance committee, committee seniority and party seniority. Column 1 presents the estimates for all bills using a linear probability model. Column 2 presents the estimates for bills loosening financial regulation using a linear probability model. Column 3 presents the estimates for bills tightening financial regulation using a linear probability model. Robust standard errors are presented in parentheses. Table 6: Regressions Results - All campaign contributions except from the Financial Sector (1) (2) (3) All All looser All tighter b/se b/se b/se Ideology 0.153*** *** (0.01) (0.01) (0.02) Total Campaign -1.02e e e-10 Contributions from Outside of Finance in the Last Election Cycle (0.00) (0.00) (0.00) Real GDP growth *** *** (0.00) (0.00) (0.00) Fraction Population working in Finance, Insurance and Real Estate (0.27) (0.28) (0.36) Constant 0.487*** 0.841*** 0.846*** (0.02) (0.02) (0.03) R-sqr N A */**/*** next to coefficient indicates significance at the 10/5/1% level. 28

29 Table 7 presents the coefficients estimates relating voting patterns to different measures of campaign contributions made by financial firms. The dependent variable takes the value 1 if the lawmaker voted in favor of a bill proposing looser financial regulation or against a bill proposing tighter financial regulation, and zero otherwise. Political controls include being in a finance committee, committee seniority and party seniority. Robust standard errors are presented in parentheses. Table 7: Regressions Results - Different measures of Campaign Contributions (1) (2) (3) (4) (5) vote vote vote vote vote b/se b/se b/se b/se b/se Ideology 0.101*** 0.101*** *** *** 0.138*** (0.02) (0.02) (0.02) (0.02) (0.01) Real GDP growth *** *** *** *** *** (0.01) (0.01) (0.01) (0.01) (0.00) Fraction Population working in Finance, Insurance and Real Estate (0.48) (0.48) (0.47) (0.47) (0.27) ln(average Campaign *** Contributions from Finance in the last 3 cycles) (0.01) ln(average Campaign *** Contributions from Finance in all preceding cycles) (0.01) ln(total Campaign *** Contributions from Finance in the last 3 cycles) (0.01) ln(total Campaign *** Contributions from Finance in all preceding cycles) (0.01) Percentage of Campaign 0.360*** Contributions from Finance in the last cycle (0.02) Constant *** (0.10) (0.12) (0.11) (0.12) (0.02) R-sqr N A */**/*** next to coefficient indicates significance at the 10/5/1% level. 29

30 Table 8 looks only at bill proposing looser financial regulation. It presents the coefficients estimates relating voting patterns to different measures of campaign contributions made by financial firms. The dependent variable takes the value 1 if the lawmaker voted in favor of a bill proposing looser financial regulation and zero otherwise. Political controls include being in a finance committee, committee seniority and party seniority. Robust standard errors are presented in parentheses. Table 8: Regressions Results - Different measures of Campaign Contributions for loosening Bills (1) (2) (3) (4) (5) vote l vote l vote l vote l vote l b/se b/se b/se b/se b/se Ideology * (0.02) (0.02) (0.02) (0.02) (0.01) Real GDP growth * (0.01) (0.01) (0.01) (0.01) (0.00) Fraction Population * working in Finance, Insurance and Real Estate (0.47) (0.48) (0.47) (0.46) (0.28) ln(average Campaign *** Contributions from Finance in the last 3 cycles) (0.01) ln(average Campaign *** Contributions from Finance in all preceding cycles) (0.01) ln(total Campaign *** Contributions from Finance in the last 3 cycles) (0.01) ln(total Campaign *** Contributions from Finance in all preceding cycles) (0.01) Percentage of Campaign 0.219*** Contributions from Finance in the last cycle (0.02) Constant ** 0.827*** (0.11) (0.12) (0.11) (0.12) (0.02) R-sqr N A */**/*** next to coefficient indicates significance at the 10/5/1% level. 30

31 Bills Table 9: Bills Details Bill Name Year Details Related Legislation H.R (103rd): Resolu It provided funding for the resolution of S. 714 failed savings associations. tion Trust Corporation Completion Act H.R (103rd): Riegle 1994 It reduced administrative requirements for Community Development and Regulatory Improvement Act of 1994 H.R (103rd): Riegle- insured depository institutions to the extent consistent with safe and sound banking practices, to facilitate the establishment of community development financial institutions It relaxed the interstate restrictions on Neal Interstate Banking and Branching Efficiency Act of 1994 branch banking by allowing interstate mergers between banks, subject to concentration limits, state laws and the Community Reinvestment Act (CRA) evaluations. Continued on Next Page. 31

32 Table 9: Bills Details Bill Name Year Details Related Legislation H.R (104th): National Securities Markets Improvement Act of It relaxed the Federal securities laws and the Investment Company Act of 1940 through exempting national securities exchange members, brokers and dealers from federal margin requirements and repealing borrowing and lending restrictions imposed upon these same institutions. H.R. 10 (105th): Financial 1998 It relaxed provisions regarding the affil- H.R. 10 Services Act of 1998 iation among securities firms, insurance (106th) companies and depository institutions. & S. 900 (106th) S. 900 (106th): Gramm- Leach-Bliley Act H.R (106th): Commodity Futures Modernization Act of 2000 Continued on Next Page It repealed part of the Glass-Steagall Act and allowed financial institutions to act as any combination of commercial bank, investment bank and insurance company It excluded several financial products from coverage in the Commodity Exchange Act, including foreign currency, government securities, security warrants, mortgages and mortgage purchase commitments. H.R

33 Table 9: Bills Details Bill Name Year Details Related Legislation H.R (107th): Financial Services Antifraud Network Act of 2001 H.R (107th): Sarbanes- Oxley Act of It tried to streamline and facilitate the antifraud information-sharing efforts of Federal and State regulators It prohibited an auditor from performing non-audit services contemporaneously with an audit; prohibited personal loans extended by a corporation to its executives and directors; required that annual reports to include an internal control report. S & H.R H.R. 314 (108th): Mortgage Servicing Clarification Act 2003 It amended the Fair Debt Collection Practices Act to exempt mortgage servicers H.R (108th): from certain requirements of the Act with respect to federally related mortgage loans secured by a first lien It allowed the ratio of reserves against its Financial Services Regulatory Relief Act of 2004 transaction accounts to be zero for depositary institutions. It also loosened the rules on operations of foreign banks in the U.S. Continued on Next Page. 33

34 Table 9: Bills Details Bill Name Year Details Related Legislation H.R. 923 (108th): Premier 2004 It amended the Small Business Investment Certified Lenders Program Improvement Act of 2004 Act of 1958 to allow certain premier certified lenders to elect to maintain an alternative loss reserve. H.R (109th): Federal 2005 It required housing-related Government- H.R Housing Finance Reform Act of 2005 Continued on Next Page. sponsored enterprises to establish an affordable housing fund to (1) increase homeownership for extremely low- and very low-income families, (2) increase investment in housing in low-income areas and areas designated as qualified census tracts or an area of chronic economic distress; (3) increase and preserve the supply of rental and owner-occupied housing for extremely low- and very low-income families; and (4) increase investment in economic and community development in economically underserved areas. (110th) 34

35 Table 9: Bills Details Bill Name Year Details Related Legislation S (109th): Financial Services Regulatory Relief Act of 2006 H.R (109th): FHA Manufactured Housing Loan Modernization Act of 2006 H.R. 698 (110th): Industrial Bank Holding Company Act of It allowed the Fed to pay interest on certain reserve balances of depositary banks. It also loosened the regulation on savings and loans departments It relaxed the manufactured housing loan insurance program under title I of the National Housing Act It strengthened the powers of the regulators to supervise Industrial Bank Holding Companies and limited commercial ownership of these companies. H.R H.R (110th): Expand It relaxed the requirements to acquire a S ing American Homeownership Act of 2007 H.R (110th): Mortgage Reform and Anti-Predatory Lending Act of 2007 single family home for underserved borrowers It established licensing and registration requirements for residential mortgage originators and provided minimum standards for consumer mortgage loans. Continued on Next Page. 35

36 Table 9: Bills Details Bill Name Year Details Related Legislation H.R (110th): 2008 It gave the U.S. treasury Department up Emergency Economic Stabilization Act of 2008 to $700 billion to recapitalize the banking sector by direct acquisition of distressed H.R (110th): Housing mortgage backed securities or new equity. and Economic Recovery Act of It authorized the Federal Housing Administration to guarantee up to $300 billion, injecting capital into Fannie Mae and Freddie Mac and authorized States to refinance subprime loans using mortgage revenue bonds. H.R (110th): Economic Stimulus Act of It provided economic stimulus through recovery rebates to individuals, incentives for business investment, and increased FHA loan limits. H.R (110th): Commodity Markets Transparency and Accountability Act of It required parties in commodities trading to provide greater amounts of information on their positions, as well as, it required the Commodity Futures Trading Commission to public provide more of that information. Continued on Next Page. 36

37 Table 9: Bills Details Bill Name Year Details Related Legislation H.R (111th): To amend the executive compensation provisions of the Emergency Economic Stabilization Act of 2008 to prohibit unreasonable 2009 It restricted financial institution that received a direct capital investment under the Troubled Asset Relief Program (TARP) from making a compensation payment to an executive or employee. and excessive compensation and compensation not based on performance standards. H.R (111th): Mortgage Reform and Anti-Predatory Lending Act 2009 It restricted some predatorial mortgage origination practices and to provided minimum standards for consumer mortgage H.R (111th): loans It provided shareholders with an advisory Corporate and Financial Institution Compensation Fairness Act of 2009 vote on executive compensation and required regulators to prescribe rules that prohibited any compensation structure or incentive-based payment arrangement that encourages inappropriate risks by financial institutions. Continued on Next Page. 37