Yesterday, June 11th, was primary day in Virginia. There were a few interesting races, such as the 24th Senate Republican primary between Senator Emmett Hanger and Tina Freitas, as well as a couple of surprising results. This morning, Andy Schmookler and I discussed these elections on 550 AM, WSVA. If you missed the show live, you can catch it here.
On March 14th of 2019, The United States Senate voted against President Trump who had earlier declared a national emergency on the U.S./Mexico border in an attempt to redirect federal appropriations toward the building of a southern border wall. The vote was 59 to 41 and included every Democratic member of the body along with 12 Republicans. (Cochrane & Thrush 2019). The previous day, the Senate voted against another one of the President’s stated positions, this time in an effort to stop U.S. military assistance to Saudi Arabia in their ongoing conflict with Yemen. Here the vote was 54 to 46 with 7 Republicans joining all of their Democratic colleagues. (Sanders 2019). This Yemen bill had previously passed the Senate in December of 2018, but it never came before the House due to opposition by then-Speaker Paul Ryan (Detrow 2018). When comparing the lists of individual Senators who opposed the president’s wishes in the recent votes, one finds complete overlap of Democratic Senators, but also some commonalities in the list of Republicans. They are Senators Lisa Murkowski of Alaska, Jerry Moran of Kansas, Rand Paul of Kentucky, Susan Collins of Maine, and Mike Lee of Utah (Edmondson 2019 & Pramuk 2019).
Although these two high-profile cases may not be representative of Congressional support for the president in the aggregate, it does raise an important question. How can one go about creating a model which accurately predicts Congressional vote support for the president?
Trump Vote Percentages in the 2016 Elections
Earlier in 2019, I found an ongoing project on Nate Silver’s FiveThirtyEight website which tracks how often individual members of Congress support President Donald Trump’s position on legislation. The website uses only a few variables. They are: how often a senator or representative votes the same way as the president would prefer, Trump’s two-party vote share in the 2016 elections in a given state or Congressional district, how often a member of Congress is predicted to vote with the president using this 2016 electoral data, and lastly the margin between the actual vote percentage and the one anticipated using this one variable. Why would a researcher consider this model? “Legislators who face a choice between supporting the government (and their parties) or the specific interests of their constituencies will tend to prefer the latter because, in so doing, they maximize their chances of re-election without imposing any costs on the government” (Cheibub 2009, 120). Furthermore, “if voters connect their votes in executive and legislative elections, the legislators will have incentives to support the executive on some key votes” (Cheibub 2009, 122-123). Thus, as the argument goes, a legislator should support or oppose the president through his or her votes in Congress in roughly equivalent amounts as their constituents rewarded the president with their vote in the previous election.
On the surface, it appears that this model has little predictive power. For example, as of March 15th, 2019, only 16 of the 100 Senators presently serving have actual Trump vote scores which are within 2.5% of their predicted scores. Expanding to 5 points either above or below still encompasses only 36 senators, with a majority still lying outside of this range. The largest differences are Senator Rick Scott (R-FL) who has a Trump support score which is 55.1 points higher than predicted and Senator John Tester (D-MT) whose Trump support score is 49.6 points lower than predicted. In the House of Representatives, the margins are even greater. While 136 of the 432 current members are within this 2.5-point range, at the extremes one can find Rep. David Valadao (R-CA-21) at 59.1 points higher than expected and Rep. Anthony Brindisi (D-NY-22) at 74.9 points lower than predicted. (Bycoffe 2019). Given the significant variation and the fact that over 66% of Congressional legislators fall more than 2.5 points outside of their expected values, one might make the claim that, by itself, the 2016 vote margin for President Trump is a poor predictor for levels of Congressional support.
However, if we consider our earlier list of Republican Senators, we find that four of the five of them, Murkowski, Lee, Collins, and Paul, are clustered toward the bottom of Republicans when it comes to how often their votes lineup with President Trump’s positions.
The Importance of Partisanship and Polarization
It seems obvious that partisanship is a key defining factor in all aspects of American political behavior in the present day. It would be easy to say that Congressional support for the president is driven first and foremost by partisan considerations and if this were the sole consideration of this paper, it would add nothing to the existing literature.  After all, just a cursory glance of the support score data provides ample evidence. In the U.S. Senate, even the Republican who has the lowest support score for President Trump, newly elected Mitt Romney of Utah, has a higher support score of 70% than any Democratic senator currently serving in that body. His closest cross-party competitor is, not surprisingly, Joe Manchin III of West Virginia at 58.5%. Perhaps coming as a shock though, Senator Romney supported the president on both of these high-profile rebukes mentioned in the previous section of this paper. Likewise, in the House of Representatives of the current serving members, Justin Amash from the 3rd district of Michigan has the lowest current support score for his party’s president among Republicans at 60%. Nevertheless, Amash’s support score is still higher than every single Democratic member of the House (Bycoffe 2019).
However, this clean party break is a fairly recent phenomenon. For example, looking back at support scores for President Obama during the 2009 and 2010 sessions reveals at least some level of party crossover. Among Republican Senators the Democratic president’s top three supporters in both years were Senators Collins of Maine, Snowe of Maine, and Voinovich of Ohio. Only Senator Collins remains in office as the last of the New England Republicans; Voinovich retired in 2011 and Snowe retired in 2013. Considering Democrats, in 2009 Senators Bayh of Indiana, McCaskill of Missouri, Feingold of Wisconsin, and Nelson of Nebraska expressed the greatest levels of opposition. It should be noted Republican Collins supported the president at higher levels than the Democrat Bayh. For 2010 Democrats with the highest levels of opposition, we find Senators Nelson again, followed by Feingold, and the Lincoln of Arkansas. As to their fates, Bayh retired in 2011, McCaskill lost to a Republican in 2018, Feingold lost to a Republican in 2010, Nelson retired in 2013, and Lincoln was defeated in the 2010 elections (CQ Almanac. 2009 & 2010).
Moving over to the House in 2009, a multitude of Republicans supported the president to a greater extent than Democratic Representatives Taylor of Mississippi and Bright of Alabama. Republican Representatives Cao of Louisiana and LoBiondo of New Jersey had support scores of over 66%. In 2010, both Taylor and Bright had the lowest support scores for Obama among Democrats while Republicans Cao and Representative Castle of Delaware has support scores of over 60%. Much like the case with the Senators, Representatives Taylor, Bright, and Cao lost their reelection bids to the nominee of the opposite party. Representative Castle ran for the Senate in 2010 and lost his party’s nomination. Of the four partisan contrarians, only LoBiondo continued to serve in elected office after 2010 (CQ Almanac. 2009 & 2010).
When considering the average support scores for the president in the U.S. Senate, according to the 2010 CQ Almanac, here’s what we find.
Although support from the opposition party in the U.S. Senate has remained relatively stable over time, we observe a widening gulf in presidential support by party as the average level of support among the president’s fellow partisans has been increasing.
By comparison, again using data from the 2010 CQ Almanac, the disparity of support for the president in the House between his own party and the opposition has become even more pronounced. Not only is support among the president’s own party increasing, as is the case in the U.S. Senate, but since the Carter administration, the president has been less successful at persuading opposition party members to vote for his proposals. As Therianault finds, “since the early 1970’s, the Senate has polarized about 80 percent as much as the House” (Theriault 2008, 197).
Not only has partisanship played a role in predicting presidential support scores in the past, but partisanship is also becoming increasingly an even more important indicator as polarization in both the House and Senate expands.
While in an earlier era, it may have been possible for scholars accurately to assert that political parties were of little theoretical importance in explaining political behavior and legislative results in the House, it is certainly not true now. Parties are consequential in shaping members’ preferences, the character of the issues on the agenda, the nature of legislative alternatives, and ultimate political outcomes, and they will remain important as long as the underlying forces that created this partisan resurgence persist (Rohde 1991, 192)
The 2016 presidential elections continued the longstanding gender gap trend in American politics. According to the Pew Research Center, women preferred Clinton to Trump by a 12-point margin. In addition, that election featured the largest gender gap since at least the 1972 election (Tyson et al. 2016). That news isn’t particularly shocking, especially given the vulgar and objectifying comments Donald Trump expressed regarding women as part of the Access Hollywood tape (Transcript 2017). The difference of attitudes between women and men regarding the president hasn’t been limited to just his election. In mid-2018, the Cook Political Report stated that “the most recent NBC/Wall Street Journal poll finds that just 39 percent of women give Trump a favorable approval rating, compared to 58 percent who disapprove of the job he’s doing. And, among white, college-educated women…the gap is staggering-just 26 percent approve to 71 percent disapprove.” Furthermore, during that time period, time white college women voters expressed their preference for a Democratically-controlled Congress by a 25-point margin (Walter 2018). According to exit polls from the 2018 midterms, 59% of women cast a ballot for Democratic Congressional candidates while only 40% picked Republicans, arguably one important reason why the Democratic Party won control of the House of Representatives in November (Velencia 2018).
Recent research has found that “eight attitudes predict Trump support: conservative identification; support for domineering leaders; fundamentalism; prejudice against immigrant, African Americans, Muslims, and women; and pessimism about the economy” (Smith & Hanley 2018, 11-12).
Considering the theory of descriptive representation advanced by Mansbridge and others (Mansbridge 1999), which advocates that in democratic systems representative legislators should not only advance the preferences of their constituents but also share other traits such as ethnicity and gender. Presumably then, given the extreme negativity women express toward the current president as compared to men, it is reasonable to expect that female members of Congress, (along with those from immigrant families, African Americans, and Muslims) irrespective of party, ought to be less inclined to support President Trump as compared to their male counterparts. 
Career Politician Support for Trump
Throughout the 2016 election cycle, the Trump campaign focused its rhetoric on three issues or slogans, “Build the wall”, “Lock her up”, and “Drain the swamp” (Overby 2017). While being sworn in in January of 2017, now President Trump continued to rail against beltway politicians. “For too long, a small group in our nation’s capital has reaped the rewards of government while the people have bore [sic] the cost” (Hemmingway 2017). Therefore, one might expect that the longer a politician is in Congress, the less likely he or she would be to support the current president. However, I would argue that the opposite is more likely the case.
As one example, consider the rather remarkable turnaround in attitude of Senator Lindsey Graham, the three-term Senator from South Carolina who also served almost a decade in the House of Representatives. A recent article from CNN explores this transformation. Before receiving the Republican nomination for president:
Graham said this of Trump: “You know how you make America great again? Tell Donald Trump to go to hell.” And, oh yeah, Graham also called Trump a “race-baiting, xenophobic, religious bigot”.
Fast-forward to the here and now. “To every Republican, if you don’t stand behind this President, we’re not going to stand behind you,” Graham said in South Carolina recently (Cillizza 2019).
So why has Graham reversed his tactics? “While Graham’s number used to lag those of other Republicans among GOP identifiers, since he has taken up the President’s banner on most every issue, his approval among Republicans in South Carolina has steadily risen” (Cillizza 2019). Senator Lindsey Graham is not burdened either by ideology or consistency and thus serves as a perfect illustration of David Mayhew’s theory that many politicians are “single-minded reelection seekers” (Mayhew 2004, 17).
Although simply looking at rhetoric and tweets might lead one to believe that Democratic leaders such as Speaker Nancy Pelosi, who has been in the House for the last 32 years, and Senate Minority Leader Chuck Schumer, who has been in the Senate for 20 years, are bitter enemies of the president and thus would have little desire to work together, much of their mutual animosity is kabuki theater. For example, as taken from an article from late March 2019, “President Donald Trump says he wants to work with Democrats to pass legislation to rebuild U.S. infrastructure…‘They want it, I want it,’ Trump said, adding that he spoke to House Speaker Nancy Pelosi ‘the other day’ about the issue” (Breuinger 2019).
While senior members of both parties have seen presidents come and go, staying in power by working to pass legislation for the benefit of their constituents and thus bolster their reelection chances, the current crop of freshman Democrats have been a largely vocal group, several of them making headlines for bucking their own party leadership openly calling for the impeachment of the president despite opposition from Speaker Pelosi. (Perticone 2019). In addition, others have found that once in office, legislators are typically relatively stable in their voting behavior (Asher & Weisber 1978) and thus new members, who are more polarized than the generation who came before, would be less likely to support the president.
Strength of the Opposition
Based upon the assumption of Mayhew, one would expect that congressional legislators are keenly sensitive to the power of presidential opposition in his or her district. In the case of a Republican President such as Donald Trump, the greater Democratic candidates perform electorally, the less likely it would be for the member of Congress, regardless of their partisan affiliation, to support the president’s agenda. “It may also be true that legislators who are truly insecure about their political standing, or that of the president, might be more willing to base their decisions on whatever local information they do have than to make risky inferences from national trends” (Borrelli & Simmons 1993, 107).
Hypothesis 1 – Representatives from states and districts which reported lower vote totals for Donald Trump in the 2016 elections ought to have correspondingly lower levels of support for him while in Congress.
Hypothesis 2 – Female legislators as a whole ought to have lower levels of support for President Trump as compared to male legislators, regardless of their party affiliation.
Hypothesis 3 – Support for the president in Congress ought to be positively correlated with legislators’ tenure in office, thus more senior members are more likely to support the president as compared to incoming freshman.
Hypothesis 4 –Higher vote totals for Democratic Congressional candidates in the 2018 midterm elections, regardless of victory, ought to correlate with lower support scores for a Republican president for the legislator of that district.
Data Collection and Analysis
I gathered the data for my regressions from several sources. My dependent variable, the Trump support score, and my independent variable of the 2016 Trump vote margin both come from Aaron Bycoffe on the website fivethirtyeight.com which he reports was compiled using data from ProPublica, Daily Kos, the Cook Political Report, and the U.S. Senate (Bycoffe 2019). Although his data includes every senator and representative who have served in any portion of the Trump presidency, I’ve restricted my analysis to current members of Congress and thus have 100 observations for the U.S. Senate and 432 for the U.S. House of Representatives. Although earlier political scientists have wrestled with the question of what Congressional votes one should consider, such as overall support, non-unanimous support, single-vote support, or the use of key votes, no matter what method one uses, so long as it is done uniformly, the differences between the measurements are usually minor. (Edwards 1985).
My remaining independent variables, Republican, Years in the U.S. Senate/U.S. House, Female, and Percentage of the Democratic vote in the last relevant general election all come from Politico as listed on four different sections on their website (2014 Election Results Senate, 2016 Election Results: Senate, House Election Results 2018, Senate Election Results 2018).
The U.S. Senate results provide highly statistically significant evidence for the first hypothesis only, which tested the theory put forth by Aaron Bycoffe, that legislators are influenced by presidential election outcomes as illustrated by the 2016 election results. A greater percentage of the vote that President Trump captured in a state in the 2016 election is positively correlated with an increased likelihood of a U.S. Senator from that state voting with the president’s wishes. Given that the Trump margin had a range of -32.2 to positive 46.3, means that two senators who have the highest and lowest Trump margins respectively are predicted to differ in support for President Trump’s legislative proposals by about 26.4% points. As expected, the partisan variable is remarkably strong, predicting a Trump support score difference of 56.46 points and it is significant at the 99.9% level.
In addition, the percentage of the Democratic vote in the last general election had a P value of .9, thus only statistically significant at the 90% CI level, but surprisingly it had a positive coefficient thus indicating that a greater level of Democratic support in a district is related to stronger support for the president. Running the model again, with the percentage of the vote for the last Democratic candidate for Senate alongside the partisan control while excluding the other previously used variables, yields a negative coefficient for the Democratic vote, as predicted, but it is still not statistically significant.
Looking at the results for the U.S. House paints a markedly different picture. Here, we find statistically significant evidence for the first three hypotheses. Although part of the explanation could revolve around the sample size, which is more than four times as large as the previous model, research from other political scientists leads me to believe that there is more to this phenomenon than such a simple explanation. As with the Senate, the partisanship plays the largest role in predicted support scores for President Trump though it is even larger than the value predicted for the U.S. Senate. This finding coincides with the research of Sean Theriault who found that party polarization in the U.S. House of Representatives is greater than what is found in the U.S. Senate. “Since the early 1970s, the Senate has polarized about 80 percent as much as the House” (Theriault 2008, 197). In addition, almost all of Theriault’s “Gingrich Senators”, members of the Senate who previously served in the House with Newt Gingrich and are believed to be more polarized than those who have not, are no longer members of that chamber.
The House’s coefficient on the 2016 Trump vote margin is only about a third as strong as it in the Senate model, though I would suspect that part of this difference stems from the increased partisan polarization as well as state legislative efforts at gerrymandering to draw as many safe, noncompetitive districts as possible within their borders. As potential evidence of gerrymandering, we observe an even greater disparity in the 2016 Trump vote margin ranging from a staggering -88.9 to a positive 63. Thus, when considering legislators from two different House districts, one with the highest observed Trump vote margin and another from the lowest, this model would predict a support score difference of 16.1 points, holding everything else equal.
As predicted, the coefficient on the female variable in the House is negative and also statistically significant. Again, this difference could stem from the fact that there are more women in the House, so the sample size is larger. By proportion, they are roughly equal at the present time. While 25 of the 100 U.S. Senators are female or 25%, 102 of the 432 or 24% of Representatives are female. But there is a considerable disparity in partisanship between the two groups. While 32% of female Senators are Republican, less than half of that number, 13%, of women in the House are members of the GOP (Women in the U.S. House of Representatives 2019, Women in the U.S. Senate 2019).
Lastly, as was the case with the Senate, the coefficient of the 2018 Democratic vote percentage in the district is positive and this time statistically significant. Running the regression again with just the last Democratic vote tempered by partisanship still produces a positive coefficient, therefore I have to conclude that my hypothesis that greater support for Democrats in a district should produce lower Trump support scores does not hold up, at least with this data set. I would be interested to see if other researchers have found similar outcomes, and, if so, what can account for this result.
As for the remaining hypotheses, if the reader will recall, the Senate data only provides evidence only for the first hypothesis, that lower Trump margins in the 2016 election coincide with lower support scores for the president. By comparison, the House data indicates backing for the first hypothesis along with the second, that female legislators ought to be less likely to support Trump’s proposals as compared to their male counterparts, and the third, the longer a representative has been in office, the greater likelihood it is that he or she will back the current president.
When David Mayhew wrote Congress, the Electoral Connection back in 1974, he observed that when it comes to the United States Congress, “its parties are exceptionally diffuse. It is widely thought to be especially ‘strong’ among legislatures as a checker of executive power” (Mayhew 2004, 7). Although presumably true at the time that they were written, his words sound out of place in the present American political climate where many activists expect their elected officials to steadfastly stand with their party’s president or in opposition to the other party’s president regardless of supposed party principles or previously held positions. But Mayhew wrote during a period when the parties were less cohesive and before the rise of polarization in the mid-1990s. It would be interesting to hear how he would update the theories in his book if it were written today. As he admits in the preface to the second edition, published in 2004, “I have not tried to revise or update this 1974 work. That would be a nightmarish task” (Mayhew 2004, xiii).
The once common conservative Democrat, liberal Republican, or ideologically moderate Congressman has become a relic of a bygone area. Although non-conformists thrived in the mid 20th century, due to the pressures of partisan polarization, by the 1990s they had become all but extinct. Legislators such as Jim Jeffords of Vermont or Richard Shelby of Alabama who often voted against the interests of the majority of their party or their party’s president either ideologically sorted themselves into a different party or found themselves replaced by partisans who did a better job at toeing the party line (Fleisher and Bond 2004).
Although the parties have split, in part over support for the current president, given historic trends one does have to wonder about the fate of Congressional Republicans who oppose President Trump more than their fellow partisans or Democrats who unduly support him. As two examples, there is talk that Representative Amash may end up leaving the Republican Party and seeking the Libertarian nomination to challenge Trump in 2020 (Kopp 2019). Presumably, if he were to do so, he would be expelled from the party and likely lose his seat in the House should he decide to run again. On the other side of the aisle, there are rumors that Senator Manchin might run for West Virginia Governor in 2020 (Everett 2019). If successful, the Senate would lose one of the few Democrats left in an increasingly Republican state.
As mentioned in an earlier footnote, as an avenue of future exploration along the lines of descriptive representation, it would be interesting to explore additional personal attributes of members of Congress. For example, are they are immigrants to this country or the children of immigrants? For those recently arrived individuals, do those with a European background support President Trump to a greater degree than those who come from, as the president calls them, “shithole countries” (Watkins & Phillip 2018)? What about race and religion? How much of a role do these personal factors play in levels of Congressional support?
At the end
of the day, it seems obvious that political party affiliation is the most
important factor in determining the level of a legislator’s support for
President Trump, although it isn’t the only issue at play. About a decade ago, Cheibub wrote that “separation
of power leads to independent legislators who act on the basis of their
individual electoral needs; in response to these needs, they build personal
ties with their constituencies.
Consequently, parties will play smaller roles and legislative behavior
will be more individualistic.” (Cheibub 2009, 127). But, after observing trends, especially now, during
the years of The Donald, the reverse may be the case in the United States.
“2014 Election Results Senate: Map by State, Live Midterm Voting Updates.” POLITICO. https://www.politico.com/2014-election/general/results/map/senate (April 13, 2019).
“2016 Election Results: Senate Live Map by State, Real-Time Voting Updates.” Election Hub. https://www.politico.com/mapdata-2016/2016-election/results/map/senate/ (April 13, 2019).
Asher, Herbert B., and Herbert F. Weisberg. 1978. “Voting Change in Congress: Some Dynamic Perspectives on an Evolutionary Process.” American Journal of Political Science22(2): 391–425.
Borrelli, Stephen A., and Grace L. Simmons. 1993. “Congressional Responsiveness to Presidential Popularity: The Electoral Context.” Political Behavior15(2): 93–112.
Breuninger, Kevin. 2019. “’They Want It, I Want It’ – Trump and Pelosi Are Still Talking about an Elusive Infrastructure Deal.” CNBC. https://www.cnbc.com/2019/03/22/trump-says-he-and-pelosi-are-still-talking-about-infrastructure-deal.html (April 21, 2019).
Bycoffe. 2019. “Tracking Congress in The Age of Trump.” FiveThirtyEight. https://projects.fivethirtyeight.com/congress-trump-score/ (March 15, 2019).
Castro, Joaquin. 2019. “H.J.Res.46 – 116th Congress (2019-2020): Relating to a National Emergency Declared by the President on February 15, 2019.” Congress.gov. https://www.congress.gov/bill/116th-congress/house-joint-resolution/46 (April 15, 2019).
Cheibub, José Antônio. 2009. Presidentialism, Parliamentarism and Democracy. New York: Cambridge University Press.
Cillizza, Chris. 2019. “Here’s Why Lindsey Graham Plays so Nice with Trump.” CNN. https://www.cnn.com/2019/03/21/politics/lindsey-graham-donald-trump-strong-support/index.html (April 20, 2019).
Cochrane, Emily, and Glenn Thrush. 2019. “Senate Rejects Trump’s Border Emergency Declaration, Setting Up First Veto.” The New York Times. https://www.nytimes.com/2019/03/14/us/politics/senate-vote-trump-national-emergency.html (April 13, 2019).
CQ Almanac. 2009. “Popular President Keeps His Caucus in Line, but the Stress Shows.” 2009 CQ Almanac: B-3–B-11.
CQ Almanac. 2010. “Legislative Success for Obama Came with Political Challenge.” 2010. 2010 CQ Almanac: B-3-B-11.
Detrow, Scott. 2018. “Senate Votes to End U.S. Support for War in Yemen, Rebuking Trump and Saudi Arabia.” NPR. https://www.npr.org/2018/12/12/676152310/senate-poised-to-vote-to-end-u-s-military-support-for-war-in-yemen (April 13, 2019).
Edmondson, Catie. 2019. “Senate Votes Again to End Aid to Saudi War in Yemen, Defying Trump.” The New York Times. https://www.nytimes.com/2019/03/13/us/politics/yemen-saudi-war-senate.html (April 13, 2019).
Edwards, George C. 1985. “Measuring Presidential Success in Congress: Alternative Approaches.” The Journal of Politics47(2): 667–85.
Everett, Burgess. 2019. “Manchin Weighs Bid for West Virginia Governor in 2020.” POLITICO. https://www.politico.com/story/2019/04/05/manchin-governor-west-virginia-1257778 (April 23, 2019).
Fleisher, Richard, and John R. Bond. 2004. “The Shrinking Middle in the US Congress.” British Journal of Political Science34(3): 429–51.
Hemingway, Mark. 2017. “On Day One, Trump Declares War on the Washington Establishment.” The Weekly Standard. https://www.weeklystandard.com/mark-hemingway/on-day-one-trump-declares-war-on-the-washington-establishment (April 14, 2019).
“House Election Results 2018: Live Midterm Map by State & Analysis.” POLITICO. https://www.politico.com/election-results/2018/house/ (April 14, 2019).
Kopp, Emily. 2019. “Rep. Justin Amash Considers Libertarian Challenge to Trump.” Roll Call. https://www.rollcall.com/news/congress/rep-justin-amash-considers-libertarian-challenge-to-trump (April 23, 2019).
Mansbridge, Jane. 1999. “Should Blacks Represent Blacks and Women Represent Women? A Contingent ‘Yes.’” The Journal of Politics61(3): 628–57.
Mayhew, David R. 2004. Congress: The Electoral Connection. New Haven, CT: Yale University Press.
Overby, Peter. 2017. “Trump’s Efforts To ‘Drain the Swamp’ Lagging Behind His Campaign Rhetoric.” NPR. https://www.npr.org/2017/04/26/525551816/trumps-efforts-to-drain-the-swamp-lagging-behind-his-campaign-rhetoric (April 14, 2019).
Perticone, Joe. 2019. “Freshman Rep. Rashida Tlaib Bucks Pelosi, Urges Democrats Not to Take Trump Impeachment off the Table Even after Mueller Findings.” Business Insider. https://www.businessinsider.com/rashida-tlaib-quietly-pushes-to-impeach-trump-2019-3 (March 30, 2019).
Pramuk, Jacob. 2019. “Trump Tweets ‘VETO!’ after Senate Votes to Block His Border Emergency Declaration.” CNBC. https://www.cnbc.com/2019/03/14/senate-votes-to-block-trump-border-wall-national-emergency-declaration.html (April 13, 2019).
Rohde, David W. 1991. Parties and Leaders in the Postreform House. Chicago: University of Chicago Press.
Sanders, Bernard. 2019. “S.J.Res.7 – 116th Congress (2019-2020): A Joint Resolution to Direct the Removal of United States Armed Forces from Hostilities in the Republic of Yemen That Have Not Been Authorized by Congress.” Congress.gov. https://www.congress.gov/bill/116th-congress/senate-joint-resolution/7 (April 15, 2019).
“Senate Election Results 2018: Live Midterm Map by State & Analysis.” POLITICO. https://www.politico.com/election-results/2018/senate/ (April 13, 2019).
Smith, David Norman, and Eric Hanley. 2018. “The Anger Games: Who Voted for Donald Trump in the 2016 Election, and Why?” Critical Sociology44(2): 195–212.
Theriault, Sean M. 2008. Party Polarization in Congress. Cambridge: Cambridge University Press.
“Transcript: Donald Trump’s Taped Comments About Women.” 2017. The New York Times. https://www.nytimes.com/2016/10/08/us/donald-trump-tape-transcript.html (April 14, 2019).
Tyson, Alec, Shiva Maniam, Alec Tyson, and Shiva Maniam. 2016. “Behind Trump’s Victory: Divisions by Race, Gender and Education.” Pew Research Center. https://www.pewresearch.org/fact-tank/2016/11/09/behind-trumps-victory-divisions-by-race-gender-education/ (April 14, 2019).
Velencia, Janie. 2018. “The 2018 Gender Gap Was Huge.” FiveThirtyEight. https://fivethirtyeight.com/features/the-2018-gender-gap-was-huge/ (April 14, 2019).
Walter, Amy. 2018. “Can Republicans Bridge the Gender Gap?” The Cook Political Report. https://cookpolitical.com/analysis/national/national-politics/can-republicans-bridge-gender-gap (April 14, 2019).
Watkins, Eli, and Abby Phillip. 2018. “Trump Slurs Immigrants from ‘Shithole Countries’.” CNN. https://www.cnn.com/2018/01/11/politics/immigrants-shithole-countries-trump/index.html (April 24, 2019).
“Women in the U.S. House of Representatives 2019.” 2019. CAWP. https://cawp.rutgers.edu/women-us-house-representatives-2019 (April 20, 2019).
“Women in the U.S. Senate 2019.”
2019. CAWP. https://www.cawp.rutgers.edu/women-us-senate-2019
(April 20, 2019).
 Noting the exception of political scientists in the mindset of Keith Krehbiel who have argued that political parties have no influence on legislative behavior.
 As I’m conducting final revisions on this paper, I realize that this thought may help explain why Representative Justin Amash (R-MI-3) has the lowest support score for President Trump among Republicans in the House of Representatives given that his father is a Palestinian immigrant and his mother is a Syrian immigrant. It would be a good variable to explore in future research.
Partisanship and the strength of the two major political parties have waxed and waned in the United States over the last several decades. After Ross Perot’s strong performance in the 1992 Presidential Elections, “a number of political observers-Theodore Lowi (1994) and Gordon Black and Benjamin Black (1994) among them-began to argue for the creation of a third major party, citing strong public support that they claim exists for a viable, mainstream alternative” (Collet 1996, 431-432). Almost two and a half decades later, public support for a viable third party remains high. According to a Gallup poll in September of 2017, only 37% of respondents believe that Republicans and Democrats do an adequate job while 61% claim that these parties do a poor job and a third party is needed. This mark of 61% in favor of a new party is the highest point in their reported data from 2003-2017. (Gallup 2017).
One does have to wonder that if the majority of the American public
In his famous work, An Economic Theory of Democracy, Anthony Downs explores the topic of party systems and ideologies in chapter 8. He argues that in a population where political opinions are normally distributed in a bell curve, two major political parties will emerge on each end of the ideological spectrum and, over time, they will become more and more alike as they converge toward the median voter (Downs 1957, 114-119). In the latter half of the chapter, Downs considers the topic of new political parties. In particular, he explores the creation of the Labour Party in Britain who, thanks to the expanded franchise of the working class, was able to position itself to the left of the Liberal Party to become the dominant party to rival the Conservatives. “New parties are most likely to appear and survive when there is an opportunity for them to cut off a large part of the support of an older party by sprouting up between it and its former voters” (Downs 1957, 128). If, as Downs claims, parties shift over time, and new parties either spring up to supplant the old ones or in order to shift a political party in a certain ideological direction, why have we not observed this trend in the United States?
Randall Holcombe suggests an answer. “In the Downsian model of political competition there are no barriers to entry, and suppliers of legislation are chosen from those who are competing to become suppliers based on how closely the output they propose to supply matches the desires of demanders” (Holcombe 1991, 233-234). So why then have these ballot access laws emerged? Holcombe compares incumbent legislators to a cartel and makes the claim that given that they control entry into the political marketplace, they can set thresholds, such as ballot access laws, to secure their continued monopoly on political power and keep competition at a minimum. “The coalition of incumbents uses barriers to entry to place challengers at a competitive disadvantage, which frees incumbents from having to strictly follow the desires of the voters” (Holcombe 1991, 231).
Thomas Stratmann explored the topic of ballot access restrictions about 15 years ago by examining both signature requirements and filing fees. He found “that monetary ballot access restrictions are an impediment to both major- and minor-party candidate entries into electoral contests…a $1000 increase in the filing fee leads to a 4% decrease in major-party candidates and a 43%…decrease in minor-party candidates” (Stratmann 2005, 69). He also determined that “signature requirements have no statistically significant effect on a minor-party candidate’s decision to enter the race” (Stratmann 2005, 69). I found this second statement particularly curious and one that demanded further investigation.
Unlike many other nations, the United States does not have uniform ballot access laws. “US electoral administration is decentralized and the exact voting experience (such as ballot format and voting technology) can vary from state to state (and even within states)” (Taylor 2018, 721). He goes on to add, “a cursory look at the US party and electoral systems would suggest that they represent clear-cut proof of Duverger’s law, the notion that single-seat plurality elections tend to create two-party systems” (Taylor 725). From 1946 to 2014, he reports that in the United States the effective number of political parties is 2.07 while the effective number of parliamentary parties is even lower, 1.94. (Taylor 2018, 725). He briefly touches on ballot access laws, “state election laws essentially guarantee a ballot slot for major party nominees in the United States, but not for smaller third parties (most especially not new ones)” (Taylor 2018, 727), but his central theory is that the rigid American two-party system results from a presidential system and the phenomenon of primary elections.
So, do ballot access laws make an observable difference in elections? If so, how? One particular election springs readily to mind. During Super Tuesday on March 6th, 2012, along with a multitude of other states Virginia held its Republican presidential primary. Although several candidates sought to have their names listed, including Newt Gingrich and Rick Perry, according to the Republican Party of Virginia “Gingrich and Perry fell short of the 10,000 signatures of registered voters required for a candidate’s name to be on the ballot…it was unclear if Minnesota Rep. Michele Bachmann, former Pennsylvania Sen. Rick Santorum, or former Utah Gov. Jon Huntsman submitted petitions to the state board” (NBC News). Gingrich, Huntsman, Perry, and Santorum filed a lawsuit in order to be listed as candidates but were unsuccessful (Myers 2012). As a result, Virginia voters ended up with only two choices, Mitt Romney and Ron Paul.
After this election lawmakers sprang into action to lower ballot access laws presumably in order to provide greater competition. Where primary challengers once needed 10,000 valid signatures with at least 400 signatures from each of Virginia’s 11 congressional districts to run for any statewide office, (§ 24.2-506) after 2012 both of these requirements were cut in half (§ 24.2-545). In the next election which followed these reduced requirements, 13 candidates appeared on the ballot for the 2016 Virginia Republican presidential primary. Although certainly curious, one should be wary of drawing conclusions based upon a single observation. Once more contests have taken place, it will be interesting to note if the average number of candidates on the ballot in a primary is statistically higher after the signature requirement slashing.
However, it is also important to note that unless they are running for a party’s nomination in a primary, neither Republican nor Democratic candidates need to collect any signatures to have their names placed on the ballot for a general election. But, any third party or independent candidate is still required to cross this threshold (§ 24.2-506). Along these same lines, one does have to wonder if ballot access laws are used as a tool to keep third-party and independent candidates out of competition (as Virginia State Senator Mark Obenshain (R-26) mentioned to me in late 2014).
With all of these thoughts in mind, my hypothesis is two-fold. First, there is an inverse relationship between the hurdles of ballot access and the appearance of third-party candidates. That is to say that greater restrictions on ballot access will result in fewer third-party candidates running for office and fewer restrictions will result in a greater number of third-party candidates. Second, one would expect fewer uncontested elections in times when ballot access requirements are reduced.
Testing my hypothesis turned out to be a more difficult task than originally anticipated. Unfortunately, ballot access laws are not readily available for most states and require poring through either state code or election laws. My original thought was to have “a tale of two Virginias”. Given that both West Virginia and Virginia have a bicameral General Assembly featuring a 100-member House of Delegates and either a 40- or 34-member State Senate, it seemed like a natural comparison at first glance. Virginia requires third party and independent candidates to gather 125 voter signatures to appear on the ballot for House of Delegates and 250 signatures for the State Senate while West Virginia requires 1% of the number of voters in the previous election. However, upon closer inspection, this comparison quickly became exceedingly messy. For starters, while Virginia elects its House of Delegates in every odd-numbered year and the entire Virginia Senate in every other odd-numbered year, West Virginia elects its House of Delegates in every even-numbered year along with an alternating half of the West Virginia Senate. Adding further complication, although the Virginia House of Delegates is elected in 100 single-member districts, presently the West Virginia House of Delegates is broken into 58 districts where most are elected through single-member plurality and others in multi-member block votes. Given these considerable differences in election times and district magnitude, it seemed nearly impossible to draw any meaningful comparison between the General Assemblies of the two states.
Exploring the idea of changes in ballot access further, I discovered that West Virginia had modified its ballot access laws in 1999, doubling the signature requirement from 1% of the votes cast in the previous election to 2%. However, in 2009, the West Virginia General Assembly voted to return to their earlier requirement of 1%. (West Virginia 2017). Even though I could not find the lines in the West Virginia Code which listed these changes, I contacted the office of the Secretary of State who confirmed the details mentioned above. In addition, although I would have preferred to look strictly at single-member districts, as I believe it would yield more precise results, using this information I would be able to examine the total number of third-party candidates who ran in each election cycle in years where the 1% rule was employed and others where the 2% hurdle was in place to determine if either resulted in statistically significant differences. At the same time, these results would show if times of lower ballot thresholds would yield more political competition through fewer uncontested elections.
For this data, I combed through election results from the West Virginia Secretary of State’s website. I decided to use general election data from 1996 to 2018 for several reasons. First, one of the major third parties in West Virginia, the Libertarian Party of West Virginia, was founded in 1994. Although a Libertarian candidate ran for governor in the 1980 gubernatorial election, the party did not field another candidate for this office until 1996. Similarly, the Mountain Party did not run a candidate for governor until 2000 (Historical Election Results). Second, I discovered that as one explores the publicly available data, the farther one goes back in time, the spottier information becomes. For example, for several elections in the mid-1990s, the website fails to list the party affiliation of some or all of the General Assembly candidates. Therefore, I would need to exclude such elections and thus each year would not have the same number of elections, 100 for the House of Delegates and 17 for the State Senate. Along these same lines, for uniformity purposes, I excluded the results for any special elections that took place for either of these bodies during the time frame.
Figure 1 shows both the total number of 3rd party candidates (those who are neither Republican nor Democrat) who ran for the West Virginia House of Delegates in a given year, as well as the total number of unopposed elections. As mentioned, after the 1998 elections, the General Assembly doubled the signature requirements from 1% to 2% and, as the chart shows, one does not see the number of 3rd party candidates return or exceed the 1998 levels until 2012 after the ballot requirements were lowered once more.
Running a regression with the dichotomous variable of ballot reform (the elections of 1996-1998 and 2010-2018 as opposed to 2000-2008), we find a fairly substantial coefficient of 6.14 with a P value of .05, lending considerable support for the idea that yes, in the West Virginia House of Delegates 3rd party competition is statistically higher when ballot access laws are at a lower threshold.
By comparison, running a similar regression using the same dichotomous variable of ballot reform, we find that, on average, that during times of lower ballot access thresholds, there are about 4 fewer uncontested races per election cycle. However, this result is not statistically significant.
Moving on to the West Virginia State Senate we observe a similar pattern. Right before the ballot access signature requirement was doubled, there was a high-water mark of 3 3rd party candidates that
Looking at a regression for the Senate using the same variables as the previous tables, we find that this lower threshold for ballot access predicts about 1.5 more 3rd party candidates per cycle as compared to the higher percentage. The P value is .09, which is higher than the desired .05 or lower, but given that only 17 seats are elected each cycle and only 10 elections are in the data set, I think it still shows some important significance.
Moving on to unopposed Senate elections we find that lower ballot access laws predict 1.5 fewer uncontested contests. However, like the House results, the high P value means that we shouldn’t think too much of this outcome.
After presenting these findings in my class, Elections and Political Parties Around the World, my professor, Dr. Herron, suggested examining the results of local elections in West Virginia as well, based upon the assumption that third parties are more likely to run candidates at the local level given that costs of running such a campaign would be lower as would be the signature requirements necessary to achieve ballot access. The three largest cities in West Virginia, Charleston, Morgantown, and Huntington, seemed to be the most logical places to examine. After some preliminary research, I discovered that cities in West Virginia hold their council elections in the spring of odd-numbered years. However, with the exception of the road bond issue in 2017 and a vote on an amendment to the West Virginia Constitution in 2005, the Secretary of State’s website had no other information about odd-numbered year elections. Calling Mac Warner’s office revealed that state law did not require localities to present their municipal election results to the secretary and so they chose not to do so.
Delving further, I tried calling the clerk in the respective cities in the hopes of gleaning information. In Charleston, they told me that they did not keep any records of their previous council elections. In Morgantown, the elections are nonpartisan and thus it would be nearly impossible to determine if any of the candidates were associated with third parties. Lastly, in Huntington, the clerk told me that they did not have any electronic records of their city council elections. However, as the election results are listed in the council meeting following the election, one could go through the council minutes for the past two decades to compile the data. Unfortunately, that data collection would have to take place by hand in the municipal building in Huntington which although may prove valuable, will require considerably more time to collect. By comparison, the Virginia Board of Elections website includes results for city council elections from 2000 to the present, General Assembly elections from 1947 to the present, and federal elections all the way back to 1789 (though the data appears to be incomplete prior to 1851) (Virginia Elections Database). If I could offer a policy recommendation for the West Virginia General Assembly, I would suggest that not only should they require localities to report their results to the secretary of state, they should make certain that the records listed on the website are as complete and accurate as possible.
In a 2012 article comparing the single-member plurality systems in the United States, the United Kingdom, and India, Patrick Dunleavy declares “Duverger’s law is now junk”, finding that “perfect two-party systems like this are now found almost nowhere outside of the USA…in particular, all the major Westminster system countries have shown strong trends toward multi-
What is the difference which can account for the abundance and success of third parties in other first past the post systems and their almost nonexistence in American politics? Could stringent and unfairly applied ballot access, as mentioned in the case of both Virginia and West Virginia be the key? Although he explores a much earlier time period than the focus of this paper, Winger argues “ballot access laws for new and small political parties were lenient in the United States, prior to 1930. In fact, there were no ballot access laws in the U.S. before 1888.” As a result, “there were several substantial third parties before 1930” (Winger 1996, 322). He goes on to state that changes were made in ballot access laws due to fears surrounding the Communist Party and although these stringent requirements were successful in their aims to thwart the Communists, they also successfully squelched every other third party as well. (Winger 1996).
However, as mentioned in the early paragraphs, not all researchers agree that more stringent ballot access laws have led to a decline in American third parties. For example, although Tamas and Hindman admit, as Winger does, that “ballot access laws became gradually more difficult for third-parties since the inception of the Australian ballot…[and]…there was a dramatic drop in the percent of House districts in which third-party candidates were getting onto the ballot…[however]…except in extreme cases, ballot access laws have had only a small impact on the ability of third-parties to get their House candidates onto the ballot” (Tamas & Hindman 2014, 273). It is important to note that these researchers used data from federal level elections while I focused on state-level results. It is possible, of course, that there is a discernable difference between the two and invites further explorations.
In conclusion, the topic of ballot access laws and how they can influence third-party competition and electoral competition in general is an ongoing debate between political scientists that this paper cannot resolve by itself. However, these West Virginia results provide some clues. Given the findings of Dunleavy and others, I’d like to examine ballot access laws in both the UK and India to see if they have high thresholds or laws, as found in some US states, which create uneven playing fields so that existing political parties can use their advantage to squelch competition and potential rivals. If not, this discovery would help explain why Duverger’s Law is still alive and well in this country while other researchers have discredited it when discussing single member plurality systems elsewhere. Free and fair elections are the supposed hallmark of democratic nations. After all, unequal ballot access laws clearly violate the international concept of free and fair elections. “All candidates, parties, and political organizations that wish to run for office should be able to do so and to compete on the basis of equal and impartial treatment under the law and by the authorities” (The OSCE 2007). If one or more political parties are using electoral laws to enforce a monopoly (or perhaps more appropriately called a duopoly) and prevent potential rivals from mounting a challenge, then the wishes of a majority of the American people for more options at the ballot box will continue to go unrealized.
Collet, Christian. 1996. “Trends: Third Parties and the Two-Party System.” Public Opinion Quarterly60(3): 431–49.Opinion Quarterly60(3): 431–49.
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: Harper Collins.
Dunleavy, Patrick. 2012. “Duverger’s Law Is a Dead Parrot. Outside the USA, First-Past-the-Post Voting Has No Tendency at All to Produce Two Party Politics.” USAPP. http://blogs.lse.ac.uk/politicsandpolicy/duvergers-law-dead-parrot-dunleavy/ (November 18, 2018).
Gallup, Inc. 2017. “Perceived Need for Third Major Party Remains High in U.S.” Gallup.com. https://news.gallup.com/poll/219953/perceived-need-third-major-party-remains-high.aspx (November 30, 2018).
“GOP: Newt Gingrich, Rick Perry Fail to Qualify for Virginia Primary.” 2011. NBCNews.com. http://www.nbcnews.com/id/45782410/ns/politics-primaries/t/gop-newt-gingrich-rick-perry-fail-qualify-virginia-primary/#.XACJyBNKg8Y (November 30, 2018).
“Historical Election Results and Turnout.” Secretary of State Mac Warner. https://sos.wv.gov/elections/Pages/HistElecResults.aspx (November 18, 2018).
Holcombe, Randall G. 1991. “Barriers to Entry and Political Competition.” Journal of Theoretical Politics3(2): 231–40
Mears, Bill. 2012. “Four GOP Candidates Fail to Make Virginia Primary Ballot, Judge Rules – CNNPolitics.” CNN. https://www.cnn.com/2012/01/13/politics/virginia-gop-primary-ballot/index.html (November 18, 2018).
Riker, William H. 1982. “The Two-Party System and Duvergers Law: An Essay on the History of Political Science.” The American Political Science Review76(4): 753–66.
Stratmann, Thomas. 2005. “Ballot Access Restrictions and Candidate Entry in Elections.” European Journal of Political Economy21(1): 59–71.
Tamas, Bernard, and Matthew Dean Hindman. 2014. “Ballot Access Laws and the Decline of American Third-Parties.” Election Law Journal: Rules, Politics, and Policy13(2): 260–76.
The OSCE/ODIHR: Election Observation Handbook. 2007. The OSCE/ODIHR: Election Observation Handbook Warsaw: ODIHR.
“Title 24.2. Elections.” § 24.2-506. Petition of qualified voters required;
“Title 24.2. Elections.” § 24.2-545. Presidential Primary. https://law.lis.virginia.gov/vacode/title24.2/chapter5/section24.2-545/ (November 18, 2018).
Taylor, Steven L. 2018. “Electoral Systems in Context: United States.” In The Oxford Handbook of Electoral Systems, New York, NY: Oxford University Press. essay, 721–39.
Tullock, Gordon. 1965. “Entry Barriers in Politics.” The American Economic Review55(1/2): 458–66.
“Virginia Elections Database » Virginia Election Statistics.” Virginia Elections Database. http://historical.elections.virginia.gov/ (November 30, 2018).
“West Virginia Legislature Passes Major Ballot Access Reform.” 2017. Libertarian Party. https://www.lp.org/news-press-releases-west-virginia-legislature-passes-major-ballot-access-reform/ (November 21, 2018).
Winger, Richard. 1996. “How Ballot Access Laws Affect the U.S. Party System.” American Review of Politics16: 321–50.
On the morning of Wednesday, February 13th, Andy Schmookler and I appeared on 550 AM, WSVA for our monthly political radio hour. The main topic of the day was the ongoing controversies with the Governor, Lt. Governor, and Attorney General of Virginia. We also briefly touched on the next potential federal government shutdown and whether President Trump would get his wall funding.
About a year ago, a friend and political activist I’ve known since 2012 contacted me about him running for State Senate in the 2019 elections. Shortly thereafter, I had a similar conversation with another friend and political activist who I’ve known for almost as long. The prospect of having two new liberty-minded individuals in the Virginia Senate was an exciting idea. The only problem was that both were in the same Senate district and both were seeking to challenge Emmett Hanger (Big Government-Augusta) for the Republican nomination. For those familiar with Virginia politics, we saw this situation play out four years ago when Dan Moxley and Marshall Pattie both sought to unseat Hanger. Given the anti-Hanger vote was split, neither was able to mount a successful challenge.
Fortunately, these two friends were able to hash things out and today, February 4th, Tina Freitas has publicly announced that she is exploring the idea of running for the Virginia Senate against Emmett Hanger in the 24th district. I would assume that most readers of this site are familiar with Tina Freitas’ husband, Delegate Nick Freitas (R-Culpeper), arguably the most liberty-friendly member of the Virginia General Assembly. Over these last several years, I have had a multitude of conversations with both of them and am pleased to say that she shares my philosophy of promoting limited government and individual liberty.
As she writes in her statement:
While I respect Sen. Hanger as a man, I strongly disagree with much of his voting record and his tendency to vote in line with the Democrat agenda on key issues. For instance, he was the vote that killed Constitutional Carry and he spent the past several years pushing for Medicaid expansion, finally ramming it through
lastsession in a budget which sent two million dollars to Planned Parenthood. This is not reflective of our respect for human life, or our defenseof Constitutional Rights here in the 24th.
We are an overwhelmingly conservative district, but unfortunately we are represented as if Hanger is ashamed of the principles which we share here in the 24th. Given that the Democrats have made their new agenda clear in recent days, Hanger’s pattern of voting with the Democrats has become exponentially more dangerous.
She adds, “I will be taking this next week to determine if there are enough people who agree with me on this point and would support me in a campaign to seek our party’s nomination.”
Reclaiming Virginia from the big government, crony capitalist, anti-freedom forces which have taken root in Richmond require strong, principled leadership and I believe electing Tina Freitas is a bold step in the right direction.
On Wednesday, January 16th, Andy Schmookler and I had our 66th hour on 550 AM, WSVA. The main topics of the day included: the ongoing federal government shutdown, is President Trump an agent of the Russian government?, and the Virginia Senate passing the Equal Rights Amendment.
John Aldrich begins Why Parties? with a bold statement from E. E. Schattschneider which states that “political parties created democracy, and…democracy is unthinkable save in terms of parties” (Aldrich 3). He goes on to add that “it is necessary to have a party system, an ongoing competition between two or more durable parties” (Aldrich 12). Throughout his first chapter, he illustrates several key concepts of democratic elections. One important feature of democracy includes free and fair competition between actors seeking elected office. Strong parties, Aldrich argues aids ambitious politicians and having two or more of them serve the public interest of preventing one faction controlling the government unchecked. (Aldrich 15-16). But are these viewpoints actually reflected in the American political system?
As the author mentions, George Washington’s addresses the issue of parties in his farewell presidential address. “I have already intimated to you the danger of parties in the State, with particular reference to the founding of them on geographical discriminations. Let me now take a more comprehensive view and warn you in the most solemn manner against the baneful effects of the spirit of party generally…The alternate domination of one faction over another, sharpened by the spirit of revenge, natural to party dissension, which in different ages and countries has perpetrated the most horrid enormities, is itself a frightful despotism. But this leads at length to a more formal and permanent despotism. The disorders and miseries which result gradually incline the minds of men to seek security and repose in the absolute power of an individual; and sooner or later the chief of some prevailing faction, more able or more fortunate than his competitors, turns this disposition to the purposes of his own elevation, on the ruins of public liberty” (Washington). It is a concern that troubled Madison as well when he wrote Federalist No. 10 nine years earlier.
Aldrich points out in his third chapter that although there was considerable instability in voting coalitions in the First Congress, the body operated without the assistance of a party system. Along these same lines, in V.O. Key’s study of the South in the late 1800s to the mid-1900s, the region was dominated by a single party, the Democratic Party. “Two-party competition would have meant the destruction of southern solidarity in national politics-in presidential elections and in the halls of Congress” (Key 8-9). In some southern states, such as Virginia, a political machine, in this case, the Byrd Organization, more or less dominated politics and thus elections were not a competitive affair nor, one could argue, democratic. This idea conjures up the idea of party bosses gathering in smoke-filled rooms in order to determine who ran the state. Elections were merely a formality, window-dressing presenting the façade of democracy. However, other southern states were a different matter. As Key illustrates in his chapter regarding Alabama, although the Democratic Party nominee for any office easily bested his opponent or opponents in the general election, Republican or otherwise, the race for the Democratic primary was often a lively affair. He shows in Table 3 that 7 candidates sought the nomination for Alabama’s 8th Congressional District in 1946. The state was divided not on the basis of party, but regionalism, with candidates typically receiving a high percentage of the vote in either their home counties, neighboring counties, and counties with which they or their campaign has some association (Key 38-43). To quote Aldrich, “until recently being a Republican in the South provided a reputation, but one that made winning all but impossible” (Aldrich 49).
Returning to the beginning of Aldrich, he writes, “the political party as a collective enterprise…provides the only means for holding elected officials accountable” (Aldrich 3). But is this statement necessarily true? Consider the case of Representative Eric Cantor (VA-7). Over time his district voters were growing dissatisfied with him. Normally re-elected with at least 63% of the vote in his previous elections, his vote percentage dropped to the high 50s in the 2010 and 2012 contests. Nevertheless, the district was a safe Republican (Sabato) and, given his influence as the House Majority Leader, the party leadership had no incentive to replace him. With Downs’ median voter theorem, parties will seek to converge toward the ideology of the largest segment of segment of the population (Downs) but Cook rated the 7th as Solid Republican (Ballotpedia) so it would be difficult for a Democratic candidate to position him or herself so far right on the ideological spectrum to mount a serious challenge, especially against a powerful incumbent like Cantor. Therefore, citizens had little chance to hold him accountable in a general election given the makeup of the district due, in part, to gerrymandering by the Republican-controlled legislature. In 2014, an unknown college professor named David Brat shocked the nation when he successfully defeated Cantor for the Republican nomination by challenging him as a more strident conservative. With this Brat/Cantor illustration, it isn’t really the two-party system holding elected officials accountable, but rather an opportune candidate seizing the right moment within a single party. Updating this example with recent data, in 2016 the courts ruled that Virginia’s 3rd district was unconstitutionally gerrymandered, therefore the surrounding districts, including the 7th, were redrawn making it more competitive. Thus, what was a safe Republican district several years ago ended up switching to the Democratic Party by a narrow margin in the 2018 elections (New York Times).
Although Aldrich might decry it as undemocratic, I would argue that this sort of state and regionalism that Alabama experienced in the early 1900s was what the writers of the Constitution expected would happen in American politics…at least before the advent of national political parties. After all, in Article II, Section 1 of the Constitution the writers seek to safeguard against states simply voting for their favored sons for both president and vice president by stating, “The Electors shall meet in their respective States, and vote by Ballot for two Persons, of whom one at least shall not be an Inhabitant of the same State with themselves.” (U.S. Constitution). We see this Alabama situation play out in the presidential election of 1824 with four candidates from the Democratic-Republican Party each winning his respective home state. As no candidate received a majority of the Electoral College vote, the decision fell to the House of Representatives. Rewinding to the previous election, known as the Era of Good Feelings, the collapse of the Federalist Party led James Monroe to an easy reelection with no serious opposition. As Aldrich writes, “the birth of party politics in a form recognizable even today can be fairly be dated to 1828” (Aldrich 102). Setting aside the issue that only a small fraction of the population was eligible to vote, despite the lack of a stable two-party system prior to 1828, I have not found much literature to suggest that the United States was not considered democratic from its founding to 1828 nor much support for the claim that political parties created democracy given that the United States government predates the party system. As another example, in the city of Staunton, Virginia, the birthplace of President Woodrow Wilson, both city council and school board elections are nonpartisan affairs. Even without the lack of party labels and cues, competitive elections are common in Staunton, and as far as I’ve found no one has declared that democracy is dead in the Queen City of the Shenandoah Valley.
Even though most elections feature candidates nominated by one or more major political parties, to argue, as Schattschenider does, that “political parties created democracy, and…democracy is unthinkable save in terms of parties” (Aldrich 3), does not appear have much supporting evidence in the history of the American political system. After all, political parties didn’t begin to take shape until many years after the founding of the present government. In addition, two-party competition was and is still absent in some regions and localities. Nevertheless, spirited competition can still persist in its absence. Democracies do not necessarily require a multitude of political parties or any parties at all, and, in the case of gerrymandering, strong parties can mute elections’ ability to reflect the peoples’ will through the redrawing district lines to make them safe or uncompetitive.
If parties are not necessary for democracy, are they still important? Here scholars disagree as well. Using data from ANES surveys, Aldrich displays a chart on page 265 which shows that as of late parties have become less important as apathy toward the parties has increased stating, “parties had become increasingly irrelevant but became at least as relevant to voters by 2008 as in the 1950s” (Aldrich 264). He goes on to add that “voting thus became candidate centered and parties as mechanisms for understanding candidates, campaigns, and elections became less relevant” (Aldrich 268). However, other scholars debate the theory of party decline. Exploring data from NES surveys and DW-Nominate scores, Hetherington reaches a different conclusion stating, “Although parties in the 1990s are not as central to Americans as they were in the 1950s, they are far more important today than in the 1970s and 1980s.” (Hetherington 619). Then, we have Krehbiel who looks at the partisan composition of Congress in committees and suggests, quoting David Broder, that “’the party’s over’ in the United States and perhaps winding down in Great Britain” (Kreibel 260).
Lastly, when considering their importance, how should one go about defining partisanship? Is it simply the number of voters who register to vote under the banner of a particular party? But what if these citizens don’t actually show up to vote? If they have no level of civic engagement, should they still be counted as partisans? And what about states which do not have registration by political party? Should partisanship be measured, as Hetherington does, through respondent thermometer scores of the respective parties or through straight ticket voting? However, then we run into the matter of whether feeling scores accurately reflect partisanship or could they simply be a lesser of two evils mentality? What about environments, such as Key’s observations about the solid South, where a viable candidate from a party outside of the Democratic wasn’t viable? Or how about the fact that some states offer their voters a straight ticket voting option at the very beginning of their ballots while others do not? Unfortunately, there doesn’t seem to be a single universal answer to the question of how one ought to define partisanship and the answer one selects likely plays a heavy role in determining how important partisanship is in American politics.
Returning to the question posed in the title of this paper, are political parties important and necessary for American democracy, I would argue, for the reasons mentioned, that they are not a necessity. Then are they important? They are, but their exact value is difficult to measure. Whether you love or hate them, parties provide a host of cues for voters who do not wish to expend the effort necessary to learn the details about each of the candidates running for office. And, at the end of the day, candidates who seek to win or maintain office without the support of one of the two major parties usually fail. But, as Aldrich claims in his 2nd chapter of Why Parties? political parties exist, not for the public interest, but primarily to serve the desires of ambitious politicians who seek to gain and maintain power for themselves. It is interesting to speculate what would happen in American politics if parties were to disappear overnight. If history provides an accurate guide, democracy would not be destroyed, and the causes of factions would still remain, of course, but, like the First Congress, it would be difficult to maintain two solid camps with an “us vs. them mentality”. Who can say? We might just see a more civil political environment as compared to our present state of hyperpolarization.
Aldrich, John H. 2011. Why Parties?: a Second Look. Chicago: University of Chicago Press.
Downs, Anthony. 1957. An Economic Theory of Democracy. New York: HarperCollins.
Hetherington, Marc J. 2001. “Resurgent Mass Partisanship: The Role of Elite Polarization.” American Political Science Review95(03): 619–31.
Key, V. O.  1984. Southern Politics in State and Nation. New ed. Knoxville: University of Tennessee Press.
Krehbiel, Keith. 1993. “Where’s the Party?” British Journal of Political Science23(2): 235–66.
“Sabato’s Crystal Ball.” Larry J Sabato’s Crystal Ball RSS. http://www.centerforpolitics.org/crystalball/2014-house/ (November 5, 2018).
U.S. Constitution. Art. II, Sec.1
“Virginia’s 7th Congressional District Elections, 2014.” Ballotpedia. https://ballotpedia.org/Virginia’s_7th_Congressional_District_elections,_2014 (November 6, 2018).
“Virginia’s 7th House District Election Results: Dave Brat vs. Abigail Spanberger.” The New York Times. https://www.nytimes.com/elections/results/virginia-house-district-7 (November 7, 2018).
Washington, George. Avalon Project – Washington’s Farewell Address 1796. http://avalon.law.yale.edu/18th_century/washing.asp (November 5, 2018).
This morning, December 5th, Andy and I held our monthly radio hour on 550 AM, WSVA. Our central focus today was the recent death of former President George H. W. Bush, President Trump, and the 2020 presidential elections.
On Wednesday, November 7th, Andy Schmookler and I appeared on 550 AM, WSVA to discuss the results of the 2018 elections. Due to a glitch with the software, I was unable to hear any of what Andy said during the show, so if our conversation sounds a bit disjointed, that is the reason for it.
In a break from our traditional schedule, this morning, October 29th, Andy Schmookler and I appeared on 550 AM, WSVA to discuss local, state, and national politics. Not surprisingly, the major focus of the talk centered around the 2018 elections, which will be taking place next week. We offered some predictions of outcomes as well as big issues and people which could end up swaying the results.
Our next show will be on November 7th at 9:15, the day after Election Day.