Analyzing the Legislative Productivity of Congress During the Obama Administration

Transcription

1 Western Michigan University ScholarWorks at WMU Honors Theses Lee Honors College Analyzing the Legislative Productivity of Congress During the Obama Administration Zachary Hunkins Western Michigan University, Follow this and additional works at: Part of the American Politics Commons Recommended Citation Hunkins, Zachary, "Analyzing the Legislative Productivity of Congress During the Obama Administration" (2017). Honors Theses This Honors Thesis-Open Access is brought to you for free and open access by the Lee Honors College at ScholarWorks at WMU. It has been accepted for inclusion in Honors Theses by an authorized administrator of ScholarWorks at WMU. For more information, please contact

2 Abstract Our nation looks to Congress to solve problems by producing legislation. Four congressional sessions took place during the eight years of the Obama administration; these were the 111 th ( ), 112 th ( ), 113 th ( ), and 114 th ( ). These four congressional sessions were often labeled and stigmatized as some of the least productive sessions in our nation s history. This raises the questions, what explains legislative productivity? What variables effect it? Were the four sessions that took place during the Obama administration that ineffective? How do these sessions compare to sessions that took place during past presidential administrations? This thesis will examine competing theories on what effects legislative productivity, such as polarization and divided government, and will analyze data and see what variables effected the amount of work accomplished by the four congressional sessions that took place during the Obama administration. This thesis will also examine the average amount of work done by Congress and will compare the Congress of the Obama administration to past congressional sessions, going back to the 93 rd Congress. Four variables will then be analyzed using IBM SPSS to see what affects the total amount of legislation passed by Congress and enacted into law. These four variables are the total amount of Democrats in each chamber of Congress, whether there is a divided government and/or divided Congress, and the Republican party unity voting as an average in both chambers of Congress. These four variables are being analyzed for specific reasons. The number of Democrats in each chamber is being looked at because it is hypothesized that as the number of democrats go up so will the amount of legislation produced this is because Democrats tend to be the party of bigger government and more social welfare programs, which takes more legislation to accomplish. Whether there is a divided government and/or Congress is being looked at because when divided, the two different political parties will have different legislative agendas, which would contribute to a decrease in 1

3 legislative productivity. Finally, the Republican party unity voting as an average in both chambers of Congress will be analyzed because it is hypothesized that when Republicans vote more together on average, they are doing so because their ideology is vastly different than the Democratic party s this is polarization. Additionally, when measuring Republican party unity voting, the Democrats party unity voting is also being measure. This is because when Republicans vote in solidarity, logically the Democrats must be doing the same this can be measured empirically. These four variables will be tested to see which of them correlate strongly with legislative productivity and then will be tested in a linear regression model to see which of the four either drive or slow down legislative productivity. 2

4 Our nation looks to Congress for solutions and expects results when we encounter problems. It is up to our elected officials in Congress to solve these issues by passing legislation. As simple as the idea sounds on paper, it is often not the case in practice. The ability to produce legislation in a timely manner is often what determines if a congressional session is successful and productive, or unsuccessful and ineffective. When members of both the House of Representatives and the Senate work to produce bills that reflect some sort of commonality for their constituents or certain interest groups, a congressional session is successful. High legislative output, more pieces of significant legislation produced, and smaller differences in ideology amongst members of Congress are key indicators that a congressional session was successful. An unsuccessful congressional session, on the other hand, is essentially the opposite of everything listed above: low legislative output, few pieces of significant legislation passed, and high levels of polarization amongst the members of Congress. This is being examined because it is important elected officials are sent to Congress to create solutions to problems our nation faces, not to play politically games amongst themselves. Additionally, this is being looked at because of the media, both during the Obama administration and after it concluded, there were countless articles stigmatizing these Congresses as being ineffective and unproductive. The four congressional sessions during the Obama administration have often been labeled four of the least effective Congresses in contemporary American politics. Numerous national news outlets have often criticized these sessions and have labeled them as being the worst Congress ever, the least productive in history, and unproductive (Marcos and Cox, 2014; Cillizza, 2014; Topaz, 2014). It would therefore seem that these congressional sessions should have been some of the least effective and unproductive in contemporary American politics. This raises the question, how do the congressional sessions during the Obama administration compare 3

5 to those during the Bush, Clinton, and other presidents administrations over the past 40 years? Were the four Congresses during Obamas administration that ineffective? What variables affect congressional productivity and how much is congressional productivity effected by these variables? What Explains Legislative Productivity? During the last few decades, the political arena of Congress has changed, altering the behavior of its members. These behavioral changes have had an impact on Congress s ability to pass legislation effectively. Four factors specifically have had an effect on Congress members behaviors ideological differences, party unity voting, divided government, and the size and importance of the party majority in each chamber. Differences in political ideology decrease legislative effectivity because when members of Congress beliefs on how the government should operate become so different, it creates political stalemate. Another factor that affects the behaviors of Congress is party unity voting, when the members of Congress vote in solidarity with only their own political party and hardly ever with the other party. Divided government is another element that affects behavior in Congress; this is when one party is in control of one of either the White House, the Senate, or the House of Representatives, leaving the other party in control of the other two. Which party controls a chamber and the size of its majority, meaning how many members make up the majority coalition, also affect congressional behavior. What Happened During the Obama Administration? During the Obama administration ( ), there were four congressional sessions, the 111 th ( ), 112 th ( ), 113 th ( ), and 114 th ( ). These four sessions each had their own set unique characteristics. During the time of the 111 th Congress 4

6 ( ), the then newly elected President had a unified government, meaning there was a Democrat in the White House and a Democratic majority in both the Senate and the House of Representatives, which would seem like a formula for a legislatively successful two years in the young president s first term. After the 2010 midterm elections and leading into the 112 th Congress ( ), this dynamic changed. The rise of the Tea Party saw the Republicans, led by House Majority Leader John Boehner, take back control of the House. This meant that the President faced a challenge he had not faced before, a divided government. The 113 th Congress ( ) saw the same party control in both chambers as the previous session, but was characterized by growing ideological differences between the two parties. The final congressional session during the Obama administration, the 114 th Congress ( ), saw the Democrats in the Senate lose their majority, and the ideological cleavage between the two main political parties grow even more. Polarization and Party Unity Voting It seems that the main culprit behind Congress s ability to legislate effectively is political polarization. What drives political polarization is ideology when a political party s beliefs are so strong and different that the party s members are less likely to make compromises with the other party, and party unity voting. The latter is defined by Schier and Eberly as, The percentage of House or Senate votes on which a member voted in agreement with a majority of his or her party (2016). Thomas E. Mann in American Gridlock writes in 2015 that, the parties in Congress are as polarized internally unified and distinctive from one another as at any time in history (2015). This brief sentence accurately captures the contemporary U.S. Congress. A cleavage has formed between the two main political parties in Congress, the Republicans and the Democrats, and it has grown larger in recent decades. This cleavage has caused members of 5

7 Congress to vote with their parties almost exclusively (party unity voting) and have opposing stances on nearly every key issue. Since the 1970 s, the two political parties in Congress have gone in opposite directions on the political spectrum and there has been a tremendous increase in party unity voting. The Republicans have continued to go further right, while the Democrats have pushed left. The ideological polarization of the parties is a recipe for an ineffective Congress and low legislative output. When members of Congress vote only with their own party, bipartisan compromises are less attainable, making for an ineffective Congress. Congress has become so polarized that the normal congressional proceedings are no longer performed. There is a standard procedure for how bills should go through Congress: a bill is introduced to the floor, is assigned to several committees, is debated on, and then gets a vote. In 1964, Medicare and Medicaid went through the proper congressional proceedings and passed through Congress with bipartisan support. This was during a time when the two parties of Congress were not as ideologically polarized and party unity voting was lower than today. In contrast, the passing of the Affordable Care Act in 2011 shows how deeply polarized Congress has become: the bill did not go through the proper congressional proceedings and not a single Republican voted for the bill. Another example of how polarized the two political parties have become was in 2013, when the federal government was held hostage and almost shutdown because President Obama and his Democratic party, which controlled the Senate, could not reach an agreement over the federal budget with the Speaker of the House John Boehner and the Republicans. (Schier and Eberly, 2016; Theriault, 2015; Binder, 2017). Congress has not always been this polarized. For a majority of the early 20 th century, the Democrats controlled both chambers of Congress. Republicans, during this time, were more likely to make compromises with the opposition hoping some of their legislative agenda could be 6

8 enacted. Then, in the 1950 s, this dynamic changed. During this time, there were intra-party caucuses representing different ideological wings of a party, i.e. northern liberal Democrats and southern conservative Democrats - which would often form different coalitions with the Republican Party. Because of this, the American Political Science Associations (APSA) created a report called Toward a More Responsible Two-Party System: A Report of the Committee on Political Parties. This guide stressed the two political parties in Congress should be more responsible and the importance of party unity voting supporting your party s platform regardless of which region of the country you hailed from. The reasoning behind this was if constituents in one part of the country were voting for a party, the members they were voting for should be more ideologically aligned with members of the same party throughout the country. (Schier and Eberly, 2016). As members of the parties have become more liberal and more conservative respectively, the ideological middle has shrunk to the point of almost nonexistence. Without these members, legislative agreements are harder to accomplish. Members in the middle served as a vital bridge between the two radical ends. Since these members of Congress no longer exist, polarization has contributed to legislative gridlock. (Binder, 2017; Theriault, 2015) The rise in party unity voting has also played a significant part to political polarization in Congress today. There was a significant rise in party unity voting in Congress in the 1980 s, and party unity voting has continued to grow more in recent decades. (Schier and Eberly, 2016; Bond, Fleisher, and Cohen, 2015). Polarization and Majority Control 7

9 Increased competition for a majority control the parties of Congress competing for majority control so they can be in charge of setting the agenda is another reason why Congress is so polarized today. In the book Insecure Majorities, Frances Lee argues, Competition for majority status affects strategic decisions in Congress (2016). The parties of Congress do not make any political moves, whether it being the passing of a piece of legislation or a public stance on a controversial topic, without keeping in mind how the move could affect the party s position in either gaining or keeping majority status. Both parties of Congress, the Democrats and Republicans, are guilty of this. Strategizing how a political move s outcome is part of the decision-making process that dominates the contemporary US Congress. (Lee, 2016) This increase in the competition for a majority control is done largely through party messaging, which is a strategy political parties use to get voters to vote for party A instead of party B. They do this by making it clear what their stances are on issues and how different they are from their political opponents. It is important for party members to make a stark contrast between them and the other party they want the voters to see this clearly. They do this because constituents need to be convinced the parties are so significantly different that voters take the initiative to vote, so their party can secure a majority. These differences on what either party s stance is on an issue or policy needs to be clear with no murkiness. These intra-party behaviors, making it clear what a party stands for without any ambiguity, make the parties vote almost exclusively together (party unity voting), because the differences between the parties need to be clear. This contributes to the expanding ideological cleavage between the two parties. (Lee, 2016) Party messaging and trying to smear the opposite party at all cost, seems to be more important in today s Congress than bipartisan voting. Bipartisanship no longer helps a party s 8

10 image because if a party is trying to make the differences between them and the other party clear, voting together does not illuminate this contrast. If a party can pass a piece of legislation without the help of the other party, it shows that the other party is irrelevant. If a party cannot pass a piece of legislation, it then allows that party to blame the other. Both instances are better for either party s image than bipartisanship because it reinforces their voter bases and makes these voter bases more eager to participate in the political process. (Lee, 2016) Stronger partisanship amongst voters makes those that identify with a political party more eager to vote for their party, which can help their party secure a majority. These constituents, with stronger partisanship, typically are out on the ends of the political ideology spectrum strong partisans are more liberal or more conservative, and hardly identify as being in the middle of the two ideologies. These stronger political ideologies amongst voters make the parties in Congress have stronger partisanship. (Lee, 2016) The Role of Divided Government Whether there is unified or divided government also plays a role in how effective Congress is. As stated above, unified government is when one party, either the Democrats or Republicans, controls the White House, the Senate, and the House of Representatives. Divided government is the opposite of this, when one party is in control of two out of the three above positions of power, leaving the other party in control of the other. Some examples of a divided government are when a Republican president is in the White House while Democrats control both chambers of Congress, or when a Democrat is in the White House, and the Democrats have a majority in the Senate, but the GOP holds a majority in the House of Representatives. If one party controls the Senate, while the other controls the House, it will be hard to get any significant and meaningful legislation to the president s desk for signing into law. The same would be true if 9

11 the president is of a different political party from the one that controls both chambers of Congress legislation may easily pass thru Congress, but a president of the other party will likely veto any significant legislation, at least in recent decades with how polarized Washington DC has become. The conventional thinking on whether unified or divided government produces more legislation seems to be clear: a unified government (one party running the show) should have a frictionless legislative process, and the result should be more legislation produced. However, this may not be the case. Research has found that having a unified government does not lead to more legislative accomplishments, but instead reduces the chances of legislative gridlock, a political stalemate between the two parties. Regardless if there is the government is unified or dived, roughly the same amount of legislating is accomplished (Ansolabehere, Palmer, Schneer, 2017). However, under a dived government, the minority party can contribute to making passing legislation more difficult. If there is unified government, the minority party has less power to disrupt the majority s agenda, as opposed to a divided government in which the minority party has some advantage over passing legislation (Binder, 2017; Mayhew, 2005). In a divided government, an opposing party can try to block legislation from being passed, but ultimately this does not affect the overall amount of legislation produced. (Binder, 2017). In their chapter in the book Governing in a Polarized Age, Stephen Ansolabehere, Maxwell Palmer, and Benjamin Schneer try to answer the question, what has Congress done?, in a unique way. While the three professors were teaching an undergraduate political science course on the politics of Congress, they decided to teach this course with a technique that was more of a hands-on and active learning experience for their students. To do this, they assigned each student a decade of Congressional history (220 years of Congress, 22 decades), and each students was to 10

12 study his or hers assigned decade and learn as much about it as possible. After the students conducted their research, they made a database of all the significant acts of Congress over the 220-year period. For the rest of their semester, the class used this database to try to answer the question, what has Congress done? They found that over the 220-year period, a Congressional session would produce on average 8.7 pieces of significant legislation when there was divided government, and 9.8 pieces of significant legislation when the government was unified. Every year, Congress passes hundreds of bills. These bills can range from renaming a national park to major tax reform. The latter would be considered significant legislation, it has a bigger impact on Americans then the renaming of a national park. Their findings show that the difference between having a unified or divided government on average is only one piece of significant legislation, demonstrating that having unified or divided government does not significantly increase the number of pieces of significant legislation. (Ansolabehere, Palmer, Schneer, 2017) When the 220-year period was broken up into four periods, their legislative means were different. Their data shows that when the government was divided, the 1 st -36 th ( ) Congresses on average produced 115 pieces of legislation in each session, while only six (around 5%) of those pieces of legislation were significant on average for each session. When unified, the 1 st -36 th Congresses ( ) produced on average total pieces of legislation, with only 6.62 (around 5%) pieces of significant legislation on average. Consequently, during 1 st -36 th Congresses ( ), having a divided or unified Congress did not have a significant impact on the amount of legislation produced. When divided, the 80 th th Congresses ( ) produced on average total pieces of legislation and an average of (around 2%) pieces of significant legislation. The 80 th -111 th Congresses ( ), when unified, on average produced total pieces of legislation and on average (around 3%) pieces of 11

13 significant legislation. Thus, the 80 th -111 th Congresses ( ), when divided, produced on average slightly more total pieces of legislation then when unified; this seems to go against conventional wisdom, but it was only a total more pieces of legislation. However, when unified, the average number of significant pieces of legislation during this period was much bigger when compared to the 1 st -36 th Congresses. The 80 th -111 th Congresses, on average, saw a 50% increase in the number of significant pieces of legislation when the Congress is unified versus when divided. An explanation for this could be that there is more polarization in Congress in recent times than there was in our nation s early history. (Ansolabehere, Palmer, Schneer, 2017) Data and Method The purpose of this research project is to see how the four congressional sessions (111 th th, ) during the Obama administration compare to previous sessions, specifically the 93 rd -110 th ( ), in terms of legislative productivity. Twenty-two congressional sessions (93 rd -114 th, ) were analyzed for this project. These twenty-two sessions spanned over eight different presidencies. These presidents are Richard Nixon, Gerald Ford, Jimmy Carter, Ronald Reagan, George H. W. Bush, Bill Clinton, George W. Bush, and Barack Obama (The Nixon and Ford presidencies were included as one for this project, due to the overlapping Congress between the two). The total number of pieces of legislation introduced in Congress in each session will be analyzed, as well as the total number of bills passed by Congress and the number of those bills that were enacted law. Which variables drive or slow down legislative productivity will also be analyzed to see what affects legislative productivity the most. The variables about the partisan composition of government that I will examine in this project are (all per session): the total number of Democrats and Republicans in each chamber, the 12

14 majority size and majority holder in each chamber, whether there was a divided government and a divided Congress, and the President s party. The behavioral variables looked at for this project include (all per session): the average ideology for each party in both chambers (Democrats and Republicans), the average ideology of each chamber as whole, the average difference in ideology for the two parties in each chamber, and the average party unity vote as a percentage of both parties in each chamber. Measuring Ideology and Polarization Most of the data used for this research project was collected from the Brooking Institution Vital Statistics on Congress. Their database has empirical data on the behaviors of Congress from past sessions, and it is kept up to date; their website states it was last updated in September of Included in these datasets is practically anything anyone could ask for when studying Congress. It includes, for example, data on congressional staffing, campaign finance, and the demographic characteristics of Congresses members. The information gathered from the Brooking Institutions Vital Statistics on Congress datasets for this research project include: Party Unity Scores in Congressional Voting, (percent), Average Ideological Positions of House Party Coalitions, 80th - 114th Congresses, , and Average Ideological Positions of Senate Party Coalitions, 80th - 114th Congresses, For the dataset Party Unity Scores in Congressional Voting, (percent), the Brookings Institution is measuring, the percentage of members voting with a majority of their party on party unity votes (Brookings Institute, 2017), of three coalitions in both chambers of Congress: The Democrats, The Southern Democrats, and The Republicans. The Southern Democrats coalition became more obsolete overtime and since the focus of this thesis is , only the Democrats and Republican coalitions were used. The two other datasets used for 13

15 this research project from the Brookings Institution, Average Ideological Positions of House Party Coalitions, 80th - 114th Congresses, , and Average Ideological Positions of Senate Party Coalitions, 80th - 114th Congresses, , measure the ideologies of the Democrats, Republicans, and each chamber as a whole. To measure these ideological scores, the Brookings Institution uses the Poole-Rosenthal DW-NOMINATE method, which was developed by political scientists Keith Poole and Howard Rosenthal (Brookings Institution, 2017) of voteview.com. This method uses ideological scores based on members' voting records (Brookings Institution, 2017) and is a standard way to measure ideology for members of Congress amongst political scientist today (Desilver, 2014; Brookings Institution, 2017). Political polarization should be the variable that drives legislative output the most. Regardless of whether there is unified or divided government, if ideological differences cannot be set aside to create and pass legislation, low legislative output should be observed. This variable can be measured in two ways, the difference in ideology and party unity voting. For the difference in ideology, the thinking is that when there is a minimal difference between the two parties and they are more in the center of the political spectrum, then they should be able to compromise easier and pass more legislation. Conversely, the further out on the ideological poles the two parties go, the amount of legislation should go down. In addition, the average ideology in each chamber should influence that chambers amount of legislation accomplished. If a chamber s average ideology is more liberal, then that chamber should produce more legislation. The opposite can be said if a chamber is more conservative. In general, congresspersons that are more liberal are less fearful of changes and big government, while conservative members of Congress are more likely to favor slow gradual changes and smaller government. 14

16 Party unity voting is another way of measuring political polarization. When party unity voting, as a percentage, is high, that means that members of Congress are voting almost exclusively with their own parties; when this is high, legislative output should be low. When party unity voting is low, members of Congress occasionally cross the aisle and vote with their political counterparts, which should lead to an increase in the total amount of bills passed. Measuring Majorities and Divided Government The data used for overseeing which party controlled the House of Representatives and the Senate came from various congressional records. The data used for the composition of each chamber comes from each chamber s online records. For example, the House of Representatives online records show the number of Democrats and Republicans serving during each session. After this data was collected, each chambers majority was calculated, for all the sessions. Dummy variables were then created to measure if there was a divided Congress, a divided government, and which party the president at the time belonged to, for each congressional session. The size of the majority in either chamber should have an impact on legislative productivity. When a party has a majority in a chamber, that party dictates the agenda in its chamber. With a large majority, it should be able to pass bills easier if a few members of the majority defect and vote with the minority party, the bill should still be able to be passed. With a small majority, it should make it more difficult to pass bills. Whether there is a divided Congress should also influence legislative output. A divided Congress is when a party controls only one chamber of Congress, leaving the other party in control of the other chamber. If party control of Congress is divided, this should make passing a 15

17 bill more difficult to accomplish. However, if one party controls both chambers of Congress, it should be easier to pass legislation because they will have similar agendas. The last and more controversial variable this research project will examine is how the number of Democrats in both chambers of Congress affect the overall amount of legislation produced. The Democratic Party tends to be the party of a larger federal government, and take the more liberal approach to governing. It is expected that as the number of Democrats in either chamber increases, that chamber should see an increase in productivity. The opposite should be observed when the number of Democrats in either chamber decreases. Measuring Legislative Output The number of bills introduced to a chamber s floor, passed, and enacted into law were collected from Congressional Records Daily Digest located on the US Senate s website. The digest breaks down each session s workload from the past how many bills were introduced in a congressional session, and whether those bills passed and were enacted into law. The total number of measures introduced was combined for the first and second year of each session. The same was done for the total number of measures passed and the total number of bills enacted into law. These numbers reflect both the House and the Senate workload. The data analysis was conducted with IBM SPSS statistics software. The purpose for doing this was to see which variables drive legislative productivity and which slow it down. To estimate these effects, a bivariate correlation measure was used with different variables to see which had the strongest effect on the total amount of legislation passed and enacted law. After these correlation measures were completed, a multivariate linear regression model was 16

18 Pieces of Legislation constructed to determine which of the variables correlated strongly with the total amount of legislation passed by Congress and enacted into law, and which variable mattered most. Results The graph below depicts the average pieces of legislation introduced, passed by Congress, and enacted into law by the President, during each president s administration Average Pieces of Legislation Per Congressional Session NIXON/FORD CARTER REAGAN BUSH CLINTON W BUSH OBAMA President TOTAL INTRO AVERAGE TOTAL PASSED AVERAGE TOTAL ENACTED AVERAGE This graph shows the comparison of the Congresses throughout the presidencies studied in this project. The average workloads of congressional sessions, opposed to the total workloads, were compared because it is a more accurate contrast between the sessions during a one-term and two-term presidential administration; the workload of the congressional sessions that took place during a four year span (one-term) should be around half of the workloads accomplished during an eight year span (two-term). The above graph shows that out of all the presidential 17

19 administrations examined at for this project, the Congresses during the Obama administration (111th-114th, ) had the lowest average number of bills passed by Congress and enacted into law. On average, these four congressional sessions (111th-114th, ) passed pieces of legislation and only pieces of legislation would be enacted law a success rate of around 15%, which is the lowest success rate amongst all of the presidential administrations analyzed for this project. The chart below shows the success rates of bills being passed by Congress and enacted into law by the president for all of the presidencies examined in this project. President Total Passed Average Total Enacted Average Success Rate Nixon/Ford % Carter % Reagan % Bush % Clinton % W Bush % Obama % There is also a noticeable downward trend in the average pieces of passed legislation and enacted into law; fewer pieces of legislation are both passed and signed into law. An increase in party unity voting and ideological polarization could be the cause of this downward trend. The Least Productive Congress Of all the Congresses studied for this project, the 112 th ( ) was the least productive in terms of legislative output. This Congress passed only 1744 pieces of legislation and only 283 were enacted law; the fewest of all the sessions studied. There are several reasons why this session was so unproductive. Out of all the sessions studied for this project that took 18

20 place before the 112 th Congress, none was more polarized; the difference in ideology, between the two parties, was larger during the 112 th than it was during the 19 previous sessions. In addition, the 112 th Congress was the first divided session during the Obama administration. The combination of having a divided government and members being highly polarized could be why the 112 th Congress was ineffective and unproductive. Correlation and Regression Models For this rest of this project, four models were created to analyze what affects the total amount of legislation passed and the total amount of legislation enacted into law. Each chamber has two models, one for the total amount passed and the total amount enacted. For each model, four variables were tested to see what correlated with the total amount of legislation passed and enacted into law and which of the four variables matter the most; to do this a multivariate linear regression model was used. The four variables are the total amount of Democrats in the chamber, whether there was a divided government, whether there was a divided Congress, and the average party unity voting score for the Republicans in each chamber. These four variables are being tested because it is hypothesized that they will have the strongest effect of legislative productivity. If there are more Democrats in a chamber of Congress, there should be an increase in legislative productivity, because the Democrats are the party of bigger government and more social welfare programs, which require more legislation to accomplish. If there is a divided government, a decrease in legislative output should be observed. When one party controls two of the three positions of power (The White House, the Senate, and the House of Representatives), leaving the other party in control of the other one, this should lead to a decrease in the amount of legislation passed and enacted into law, because the two parties will have different political agendas, which could cause a political stalemate. The same 19

21 Senate Republicans Party Unity Voting as an Averge (%) House Republicans Party Unity Voting as an Averge (%) reasoning applies to why there should be a decrease in legislative productivity under a divided Congress. Finally, when Republicans vote as a party more on average, a decrease in legislative output should be observed. The average party unity voting of the Republicans will serve as a polarization measure for these models because the Republican party unity voting as an average and the difference in ideology between the two parties in both chambers of Congress correlate very closely to each other. The graphs below show these relationships. House of Representatives Difference in Ideology and Republican Party Unity Voting Average House of Representatives Difference in Ideology Senate Difference in Ideology and Republican Party Unity Voting Average Senate Difference in Ideolgy 20

22 These graphs show that the differences in ideology and party unity voting both increase at the same time. These strong correlations between the difference in ideology of the two parties (in both chambers) and the Republican party unity voting (in both chambers) show that they are measuring the same thing polarization. For this reason, only one will be used in the following models, the Republican party unity voting as an average. When Republicans vote together more, they are not voting on issues with the Democrats, because there are bigger differences in ideology between the two parties. In addition, when Republicans vote together more, the Democrats are also voting together, which is why the average party unity voting of the Democrats in both chambers is not being used in these models. Senate Total Passed Correlations TOTAL PASSED DEM-S DIV GOV DIV CONGRESS PARTY VOTE REP S AVE TOTAL PASSED To measure which variables have a relationship with the total amount of legislation passed by Congress but not enacted into law, a correlation measure was made to measure these effects. The chart above shows the correlations between the previously discussed four variables in the Senate and the total amount of legislation passed by Congress. Of the four variables, only one relationship is statistically significant, the party unity voting of Republicans as an average. The Pearson correlation between the party unity voting of Republicans and the total amount of legislation passed by Congress is negative This number shows that when the Republicans 21

23 Republican Average Party Unity Voting in the Senate (%) in the Senate vote together more on average there tends to be a decrease in legislation passed in Congress. The below graph shows this trend Republican Party Unity Voting and Passed Legislation by Congressional Session Total Amount of Legislation Passed by Congressional Session This graph depicts the relationship party unity voting as an average per session has with legislative productivity. It shows that when Senate Republicans vote together more, the total amount of legislation passed by Congress decreases. The variables were then put into a multivariate linear regression model to see which of the four influence the total amount of legislation passed by Congress the most. The table below shows this linear regression model. Coefficients a Standardized Unstandardized Coefficients Coefficients Model B Std. Error Beta t Sig. 1 (Constant) DEM-S DIV GOV DIV CONGRESS PARTY VOTE REP S AVE a. Dependent Variable: TOTAL PASSED 22

24 When put together, only one of the four variables tested in this model was statistical significant. The average Republican party unity voting. This shows that even when there is an increase in the number of Democrats in the Senate, a divided government, and a divided Congress, the variable that matters most is when Republicans vote together. The above table shows that for every one-percent increase in Senate Republican party unity voting as an average, the total amount of legislation passed in Congress decreases by Of the data collected, the range of Senate Republican party unity voting as an average is 23.5% (94%-70.5%). According to this regression model, this increase in Republican party unity voting came at a cost of around 1,187 pieces of legislation (23.5*50.582). This variable is a measure of polarization, when there is a bigger difference in political ideology between the two parties, the Republicans vote more together, and Congress becomes more unproductive. Although the divided Congress variable in this model was not statistically significant, it was still substantially important according to this model having a divided Congress came at a cost of around 420 pieces of legislations. Senate Total Enacted Correlations TOTAL ENACTED DEM-S DIV GOV DIV CONGRESS PARTY VOTE REP S AVE TOTAL ENACTED To measure what affects the total amount of legislation passed by Congress and enacted into law, the same method was performed as before. The above table shows the correlations between the total amount of legislation enacted into law and the four variables used in the 23

25 Republican Average Party Unity Voting in the Senate (%) previous model. The results are similar to the previous model: only one of the four variables was statistically significant, Republican party unity voting. The correlation between party unity voting of Senate Republicans and the total amount of bills enacted into law is a strong-negative correlation, The below graph depicts this relationship. 100 Republican Party Unity Voting and Enacted Legislation by Congressional Session Total Amount of Legislation Enacted Law by Congressional Session This graph has a striking resemblance to the one depicting the total amount of legislation passed by Congress. These findings show that the more Senate Republicans vote together on average, the less likely bills passed by Congress are being enacted law. This correlation again points to polarization. Coefficients a Standardized Unstandardized Coefficients Coefficients Model B Std. Error Beta t Sig. 1 (Constant) DEM-S DIV GOV DIV CONGRESS PARTY VOTE REP S AVE a. Dependent Variable: TOTAL ENACTED 24

26 As before, the four variables were put into a multivariate regression model to see which affects legislative output the most; the above chart shows this. In this model, again only Senate Republican party unity voting is statistically significant. This data shows for every 1% increase in the party unity voting of Senate Republicans as an average, the total amount of legislation enacted law decreases by Again, the range of Senate Republican party unity voting as an average is 23.5% (94%-70.5%). According to this regression model, Republican party unity voting stopped around 423 bills from being enacted into law (23.5*18.088). This regression model shows when there is a divided government, divided Congress, and more Democrats in the Senate, what effects legislation being signed into law the most is when the Senate Republicans vote together. This is an effect of polarization. When members of the Senate become more polarized, (they vote more within their parties) and legislative productivity decreases. In the House of Representatives, a similar trend should be observed. House of Representatives Total Passed Correlations TOTAL PASSED DEM-H DIV GOV DIV CONGRESS PARTY VOTE REP H AVE TOTAL PASSED The chart above shows the correlations between the total amount of legislation passed by Congress and the four variables used in the previous models: the total amount of Democrats in the House, whether there is a divided Congress or a divided government, and the party unity 25

27 Number of Democrats in the House of Representatives voting of House Republicans as an average. Of the four, two correlations are statistically significant, one is marginally significant, and the remaining not significant at all. The two statistically significant variables are the total amount of Democrats in the House and the party unity voting of House Republicans. The correlation between the total amount of Democrats in the House and the total amount of legislation passed by Congress is a positive one when there are more Democrats in the House, Congress tends to pass more legislation. The chart below depicts this trend Number of Democrats in the House of Representatives and the Total amount of Legislation Passed by Congress Total Amount of Legislation Passed by Congress The other statistically significant correlation in this model, the total amount of legislation passed by Congress and the party unity voting of House Republicans as an average, is negative. This means when House Republicans vote more together on average the total amount of legislation passed by Congress tends to decrease. The graph below depicts this correlation. 26

28 Republican Averaerage PArty Unity Voting in the House of Representatives (%) Republican Party unity Voting and Passed Legislation Passed by Congressional Session Total Amount of Legislation Passed by Congressional Session This relationship is similar to the trend observed in the Senate. This is another instance of polarization affecting legislative productivity. Coefficients a Standardized Unstandardized Coefficients Coefficients Model B Std. Error Beta t Sig. 1 (Constant) DEM-H DIV GOV DIV CONGRESS PARTY VOTE REP H AVE a. Dependent Variable: TOTAL PASSED As before, a multivariate linear regression model was created to see which of the four variables matters the most in terms of the total amount of legislation passed by Congress. The above chart shows the results of this model. Only one of the four variables is marginally significant (regression is significant at the.05 level), whether there is a divided Congress. This 27

29 data shows that when divided, the total amount of legislation passed decreases by in Congress. Although there is a significant correlation between the total amount of legislation passed by Congress and the number of Democrats in the House and the party unity voting of House Republicans, neither is what drives legislative productivity in the House of Representatives. According to this data, a divided Congress has more of an effect on legislation being passed. House of Representatives Total Enacted Correlations TOTAL ENACTED DEM-H DIV GOV DIV CONGRESS PARTY VOTE REP H AVE TOTAL ENACTED This table shows the correlations between the four House variables and the total amount of legislation enacted into law. The two variables that are statistically significant, the number of Democrats in the House and the party unity voting average of House Republicans, are the same two variables that were statistically significant in the previous correlation measure. However, in this model the correlations are stronger than in the previous. This means when there are more Democrats in the House of Representatives, more legislation in being enacted into law and when Republicans in the House are voting more together, fewer pieces of legislation are being enacted into law when compared to the previous correlation measure. The below graphs show these trends. 28

30 Party Unity Voting of House Republicans as an Average (%) Number of Democrats in The House of Representatives Number of Democrats in the House of Representatives and Legislation Enacted Law Legislation Enacted Law Party Unity Voting of House Republicans and Legislation Enacted Law Legislation Enacted Law There are noticeable trends occurring in this model. It is clear that one variable correlates positively with legislation being enacted law, while the other correlation is negative. These findings show that House Democrats link closely to legislative productivity while House Republicans voting together is linked to ineffective legislating, confirming the hypothesis that when there are more Democrats in a chamber, legislative productivity goes up and when Republicans vote more together on average the opposite happens. 29

31 Coefficients a Standardized Unstandardized Coefficients Coefficients Model B Std. Error Beta t Sig. 1 (Constant) DEM-H DIV GOV DIV CONGRESS PARTY VOTE REP H AVE a. Dependent Variable: TOTAL ENACTED The above regression model shows which House variables matter for the total amount of legislation enacted into law. According to this data, only the party unity voting of House Republicans matters, because it is the only variable that is statistically significant. This regression model shows that for every 1% increase in the party unity voting of House Republicans as an average, the total amount of legislation enacted into law decreases by Of the data collected, the range for House Republicans party unity voting as a percentage is 22.5 (95%-72.5%). According to this data, this increase in Republican party unity voting resulted in a loss of around 343 (22.5*15.227) pieces of legislation enacted into law. These findings again show what effect polarization has on legislative productivity. Conclusion According to the data analyzed in this project, polarization is what matters most for legislative productivity when polarized, Congress becomes less productive, and when not as polarized, the opposite trend happens. The polarization variable used in the four models, the Senate Republican party unity voting, has the strongest correlation for legislation passed and enacted into law in three of the four models. Additionally, this variable was statistically 30