Polling Explanations and Final Takeaways from the Kentucky Senate Race

Last week, Kentucky made national headlines as Senator Mitch McConnell easily swept by his Democratic challenger  Secretary of State Alison Lundergan Grimes. While polls predicted the race to be close in the week leading up to the election, showing a 5-9 point lead for McConnell, he ended up with a 15-point win on election night. In this blog post, I will explore why there was such the discrepancy in predicted polls and the margin of victory as well as see what we learned from the 2014 Kentucky Senate election.

Why Did Polls Report a Close Race when There Was a 15-Point Margin of Victory?

According to Andrew Gelman and Gary King, when opinion-based polls circulate, people tend to respond with “enlightened answers.” Enlightened answers, to them, are not rational. Especially with polls in the beginning stages, people are responding with imperfect information—they do not exactly know what either candidate stands for, but instead make assumptions and do not take their uncertainty into account when responding to opinion-based polls. Thus, especially in the early stages, polls are inaccurate predictors of rational voting behavior (see here and here for information about poll accuracy). This, for Gelman and King, would explain the horserace of opinion polls throughout campaigns—individuals are continually gaining more and more information and thus their answers are becoming more rational as time progresses.

Was that the case with McConnell and Grimes? Looking at the aggregate polling data, as time progressed, Grimes was less and less likely to be leading in the polls. In the month leading to the election, Grimes polled ahead of McConnell a single time, and even then, it was well within the margin of error. Looking at a time period between May 2013 and June 2014, Grimes polled ahead of McConnell 9 times, and was tied with McConnell one time. Thus, most of the data used in the horserace media culture presenting the race as a tight one came from the beginning months of the election. In contrast, the one time Grimes polled ahead of McConnell during the last month of the election, it was within the margin of error, which should have been a clear predictor of McConnell’s victory; the media, however, didn’t focus on that—they often gave the impression that the race was continuing to be close.

So now that we know it wasn’t as much as a horserace as the media made it out to be, why was the margin of victory higher than the predicted poll? KET asked college professors in Kentucky for their answers to this question. Responses ranged from the idea that perhaps undecided voters broke out at the last minute, to the idea that conservative Republicans who had strayed away from McConnell ended up voting for him while Grimes lost traction with liberal Democrats, to the idea that conducting polls is harder nowadays due to caller ID and cell phones even though polling methodology has improved. In a separate interview, Centre College’s Professor of Politics Benjamin Knoll says that whatever the cause for McConnell’s over-performance relative to the polling averages, it was not limited exclusively to Kentucky. “Polling averages tended to underestimate Republican advantages in several Senate elections this year,” he said, “and thus Kentucky’s outcome fits into a wider national pattern.” Thus, there are apparently many reasons why the polls may not have matched the margin of victory.

Takeaways from the Kentucky Senate Race:

Grimes fell severely short in places where she should have won: on election night Grimes lost half of the ten most Democratic counties by voter registration—all of which are in Eastern Kentucky (see here for statistics on turnout by county)

Counties with the highest turnout were not turning out for Grimes: of the 39 counties that saw turnout to be above 50%, Grimes only won two (see here for statistics on turnout by county)

Obama helped McConnell win votes, but voters turned out for reasons other than to vote against Obama: One of every four voters said they were angered with the Obama administration; further, of those voters, ninety percent voted for McConnell; however, 46% of Kentuckians said Obama had no part in their reason for showing up to vote

Kentucky thinks the country is on the wrong track: 69% of Kentuckians think the country is on the wrong track; interestingly, around the same percentage said the same thing 6 years ago in McConnell’s previous election

Most Kentuckians already knew for whom they were voting: when asked when voters had made up their minds, 70% of voters had made up their minds more than a month before the election; interestingly, 10% of voters said they made up their minds in the three days leading up to the race, with an even distribution between those deciding to vote for any of the three candidates (all less than 1% of the candidates’ total vote percentage)

Last, but not least—41% of Kentuckians claimed they were not contacted by either Grimes’ campaign or McConnell’s: let’s say this together… “You lucky dogs!”

Why Did McConnell Win and Grimes Lose?

A little over a week after the reelection of Senator Mitch McConnell people might be left wondering what factors may have contributed to this result. In my last blog post I hope to make sense of these results.

One major factor that contributed to McConnell’s victory was his position as the incumbent. This election cycle, more so than the post-war average, incumbents won their elections by astounding margins. Senate incumbents won 85 percent of their elections this November. We discussed previously the incumbent advantage and explained why incumbents typically win reelection, with few exceptions.

Screen Shot 2014-11-13 at 3.10.25 PM

Beyond McConnell simply being an incumbent, he is a stout Republican and created himself an image as anti-Obama. With midterm elections typically being a referendum on the sitting president and his or her party’s performance, Republicans were looking to take control both houses of the legislature and they were able to accomplish this. We also previously discussed McConnell’s benefit being in the opposite party as the president. This was only aided by Obama’s low approval rating in Kentucky.

Kentuckians had an extra incentive to reelect McConnell, outside of the stark dislike for the president within the state. At least some were aware that if McConnell was reelected and Republicans took control of the Senate that he would become the Senate Majority Leader. This leadership position is really appealing to voters with hopes of McConnell using the leadership role to institute positive change not only in the Senate but also to promote issues that are pertinent to Kentuckians.

In a previous post, we discussed who Kentuckians should elect and discussed the positives and negatives to each candidate, focusing solely on institutional considerations as opposed to political policy positions. Grimes, although a solid challenger, did not have the same political experience as her opponent and did not have nearly the same immediate potential for powerful leadership if elected. Being a Democrat certainly did not help her case in Kentucky. Voters found value in McConnell’s experience, political party, and leadership potential and these all strongly contributed to his reelection to a sixth term.

Why Rational People Do Not Vote

After Tuesday’s election, we are now able to see how many Americans turned out to vote during this election cycle. Compared to a voter turnout of 40.9 percent in the 2010 midterms, only 36.6 percent of Americans turned out to vote on Tuesday.

Anthony Downs, an American economist, argued that a citizen would vote rather than abstain if the citizen’s vote value exceeds zero, in other words, if the benefits outweigh the costs. (This is called the “rational choice theory of voting.”) For a rational, self-interested voter, the costs of voting will normally exceed the expected benefits that one might receive from the effort put into voting.

So what are some of the costs that are cited as being the reason for not voting?

The biggest variables in voter participation involve traits of individual voters, which affect costs, such as political interest or education. Among other things, accessibility and convenience issues have also been shown to have a significant impact on the number of people who turn out to vote on Election Day. With Election Day falling on a Tuesday, a number of Americans do not have the time off from work or from everyday goings on to go to the polls. Public schools being let off for the day only exacerbates the situation with parents having to either bring their child to the polls or find extra child-care.

Many people think their vote does not count, especially when it comes to Presidential elections since the vote is counted in order to determine the candidate that the states’ Electoral College candidate endorses. The belief that one’s vote does not count could be considered a “cost” and only furthers general apathy toward the political system.

Beyond the apathetic nature of Americans, many claim that they are too busy to vote. Family, work, and other life events have a tendency to get in the way of people who want to vote and interrupt their ability to do so. As was previously mentioned, Election Day being on Tuesday and the polls being open, for the most part, during work hours can cause problems in peoples’ schedules and people claim they simply do not have the time.

Although there are a number of costs that are associated with voting, it is still an important civic duty and one that eligible voters ought to be excited about. However, Kentucky stood out this election cycle among national statistics with a voter turnout projected to be about 44.2 percent statewide. In fact, this statistic creeps toward typical Presidential election turnout rates. Although the country as an entity has problems with turning out to vote, Kentuckians found value and importance in these elections and nearly half turned out to exercise their rights. In fact, Kentucky’s turnout rate on Tuesday was in the top quarter for highest turnout among all states, something to be proud of!

How Local Conditions Shape Voter Turnout


In less than 30 hours, Kentucky will have determined our next senator: either Senator Mitch McConnell or Secretary of State Alison Lundergan Grimes. Up until now, this blog has taken many issues that the candidates have raised and have tried to align them with research in political science. By now, after months of campaigning and in the final stretch, I feel like we know where MCCONNELL-LUNDERGAN-GRIMES-US-SENATE-RACEthe candidates stand, and therefore this post will not look at either candidate, but rather about the conditions for Election Day. Why? Because at the end of the night on November 4th, it doesn’t matter who how much money either candidate spent or the areas where they campaigned if the voters do not show, and certain conditions on Election Day itself have proven to impact turnout in different ways. Thus, in this post, we will look at the weather conditions and voter turnout, Kentucky polling hours, the number of polling locations and its correlation to turnout, and finally the impact exit polling has on voters. Let’s begin:

Election Day Weather and Turnout:

In the political sphere, there is conventional wisdom that Republicans should “pray for rain”. Why is that? According to Brad Gomez, Thomas Hansford, and George Krause it is because when there is inclement weather, peripheral voters are less likely to turnout to vote, and these voters have historically aligned with the Democratic Party rather than the Republican Party. Therefore, if these peripheral voters do not show due to weather, it only increases the vote share of the Republican candidate. Gomez et al., find this adage to be true, showing that in presidential elections, for every inch of rain above an area’s Election Day normal, the Republican candidate received an extra 2.5% of the vote compared to the Democratic candidate. In a tight race, that could make an impact.

So what does this mean for Kentucky? Unfortunately for McConnell, the majority of the state will not be seeing any rain tomorrow; however, not all areas of the state have a 0% chance. Parts of western Kentucky may see rain, such as in Paducah where there is a 70% chance of rain, or in some parts of northern Kentucky were the percentage of rain is close to 50%. Even so, it does not serve much of an impact of McConnell, and Grimes should be thankful as she tries to secure as many possible votes against the incumbent.

Kentucky Polling Hours and Voter Turnout

Many Kentuckians may not be aware that the hours that polls are open in Kentucky are relatively short compared to many other states—our polls close at 6 pm, while other states have their polls open to as late as 9 pm. Does this make a difference in voter turnout? Turns out, not much research has been done on this topic. Kyle Dropp explores the effect of reduced polling hours in areas such as Vermont, Minnesota, and Montana and finds that in two of the three states, there was no significant effect on a change of voting hours and turnout. However, this data looks at a single state reducing its hours, rather than states that have shorter hours compared to states that have longer hours. Thus, I examined the 2012 state-by-state voter turnout rates and the amount of hours each state was open for polling, and found that really it did not make a great difference. For example, Iowa is a state that has one of the longest polling hours in the country and was ranked 5th among voter turnout rates in 2012, while Colorado was ranked 3rd and it has the same length of polling hours as Kentucky. In most other high-voting states the length of polling hours varied. Therefore, without further research, I cannot say positively whether the length of polls being open will help or hurt voter turnout. What I can say is that in Kentucky, our voter turnout rates for midterm elections are increasing over the years. In 1998 the voter turnout rate was 23.8%, by the 2002 midterm it was 26.6%, in 2006 that number jumped to 30.8%, and then by 2010 the turnout rate increased slightly to 30.9%. One can only hope that with this year’s election, Kentuckians will show up to the polls between 6 A.M. and 6 P.M. and continue to increase our turnout rate.

Exit Polling and Voter Turnout

So why even look at polling hours if there is not any evidence which suggests that they neither help nor hurt a candidate? Because while we do not know polling hours’ direct impact, it turns out that these hours do have indirect impacts—such as those on voter turnout and exit polling. According to Seymour Sudman, exit polls can reduce voter turnouts among states. Sudman finds that in states where polling hours are open later—that is, 8 P.M. and after—there was a decrease in the amount of people going to the polls ranging from 1-5% of total voters.


In some states, factors such as rain, polling hours and exit polling could have a great impact on voter behavior, but for Kentucky, it looks like no one is likely to have an advantage due to local conditions. We are interested to see what will happen, and look forward to blogging about post-race analyses after the race is said and done. The money has been spent, now it is time to see Kentuckians take advantage of their civic duties and elect our next Senator.

How the Kentucky Senate Race Could Shape the Federal Judiciary

Dr. Lee Remington Williams is the Assistant Professor of Political Science and the Pre-Law Program Co-Director at Bellarmine University in Louisville, Kentucky. Williams secured her B.A. in History and Government from Morehead State University, her J.D. from the Brandeis School of Law at the University of Louisville, and her Ph.D. in Judicial Politics from the University of Kentucky. Currently, Williams resides in the Highlands area in Louisville with her husband Bill, and her sons Remi and Ross.

The Importance of Federal Judicial Openings

As part of the doctrine of checks and balances, the U.S. Constitution allows presidents to appoint federal judges with the “advice and consent” of the Senate.  From Marbury v. Madison to FDR’s “court-packing” plan, federal judgeships have been at the heart of many of our country’s political fights.

The political maneuvering over the make-up of the federal judiciary continues today.  The filibuster has historically been used to block federal court nominees (usually appellate nominees).  Through the years, both parties have filibustered judicial nominees they deem undesirable.  This tactical procedure has been criticized by both parties (but usually only when that party’s nominees were being filibustered).  This reluctant cycle came to an impasse last November when the Senate Democratic majority invoked the so-called “nuclear option” on certain judicial nominees, prohibiting Republican filibuster and calling for a simple majority vote.

Why do presidents and senators go to such great lengths to influence the make-up of federal judiciary?  In the aggregate, there is plenty of evidence showing how presidents can impact judicial policy through their federal court nominees, especially since those judges have lifetime appointments. The more federal judge positions presidents fill, the more lasting impact presidents have upon policy.  It is no surprise then that presidents attempt to choose judges who share the president’s views.

Senatorial Influence upon the Federal Judiciary

Over time, presidents and senators established a method of choosing desirable federal court candidates through a tradition known as “senatorial courtesy”. Senatorial courtesy refers to a tradition of deference whereby the president (and Senate) will defer to the recommendation of (usually senior) senators from the President’s party within the state when there is a judicial opening (usually district court).  For example, if there is a federal district court opening in Vermont, President Obama would defer to the recommendation of Democratic Senator Patrick Leahy, and the Senate would traditionally approve the nominee as a “courtesy.”  (Note that this usually only applies when the senator is of the same political party as the president.  However, the contentious nature of the political environment today has resulted in some strange bedfellows in order to get certain nominees through.)

What about individual nominations?  How much impact could a Senator’s recommendation for one federal district court judgeship really have upon policy?

Kentucky offers a good example of such an impact, though perhaps not the impact that the recommending Senator originally envisioned.  In 1992, a federal judgeship opened in the U.S. District Court for the Western District of Kentucky.  Pursuant to senatorial courtesy, then-President George H.W. Bush deferred to the recommendation of the senator from his own party in that state, Senator Mitch McConnell.  McConnell, the former Jefferson County (Kentucky) Judge-Executive, recommended his former Special Counsel to that office, John G. Heyburn II.

Over the years, Heyburn has garnered a reputation as a moderately conservative judge.  Until this year, most would never have accused Judge Heyburn of being a “liberal” judge.  However, that all changed on February 12, 2014, when Heyburn ruled that the Commonwealth of Kentucky must recognize same-sex marriages performed in other states.  Five months later, Heyburn extended this ruling, holding that Kentucky’s own same-sex marriage ban was unconstitutional.   In both decisions, Heyburn relied upon the U.S. Supreme Court’s decision in U.S. v. Windsor, 133 S.Ct. 2675 (2013), where the Court declared that the provision of the Defense of Marriage Act (“DOMA”) defining marriage as between a woman and a man was an unconstitutional violation of the Equal Protection and Due Process Clauses.  Judge Heyburn’s decisions have been appealed to the 6th Circuit Court of Appeals, which covers four states – Kentucky, Michigan, Ohio, and Tennessee.  In each of these states, a federal judge has ruled in favor of marriage for same-sex couples (albeit under different legal doctrines).  The 6th Circuit heard oral arguments from all of these cases in August 2014 and a decision is expected to be issued at any time.  Yet, Heyburn is not alone.  Many other lower federal court judges, relying upon Windsor, have also overturned gay marriage amendments within their own states.

On October 6, 2014, the U.S. Supreme Court denied seven cert petitions from five states, including Utah, Oklahoma, Virginia, Wisconsin, and Indiana, which sought review of rulings from the Fourth, Seventh, and Tenth Circuits striking down bans on same-sex marriage.  By leaving these lower court decisions intact, the denial had the immediate effect of legalizing gay marriage within those states.  As it stands today, gay marriage is now completely legal in 32 states.  Three other states are subject to federal precedent paving the way toward gay marriage.  In eight more states, judges have issued rulings in favor of gay marriage, but these decisions are on hold pending appeal.  In Missouri, marriages of same-sex couples legally performed in other states are respected. Because Kentucky’s fate is up in the air pending the 6th Circuit’s decision, it remains one of the 18 states that do not allow gay marriage.  If Judge Heyburn’s decision is upheld, however, Kentucky will join the majority of states allowing such.

When it comes to the senator who recommended him, Heyburn’s gay marriage rulings have not gone unnoticed.  After Heyburn’s gay marriage decisions, the Senate Conservatives Fund released an ad attacking McConnell for recommending Heyburn to the bench.  The ad stated, “Senator McConnell should admit that recommending Judge Heyburn was a mistake that hurt Kentucky….McConnell knew Judge Heyburn was not a conservative, but he promoted him anyway. Now Judge Heyburn is forcing his liberal views on Kentucky”

The 2014 Kentucky Senate Race

It is evident that federal judicial positions are important, so important that legislators and presidents have been wrangling over them since our founding.  How much impact can one Senator have upon our judiciary and resulting judicial doctrines?  Think of it this way:  In each of the gay marriage cases above, many of the judges involved were put into those positions due to recommendations by senators under “senatorial courtesy.”  Long-serving senators will have more of an impact over time.  Further, if you live in a state where a senator is of the same party as the president for an extended period of time, and those presidents have the opportunity to fill numerous judicial openings, the impact of a recommending senator on judicial policy could be huge.

The 2014 Kentucky Senate race between Democrat Alison Lundergan Grimes and Republican Mitch McConnell has become one of the most watched Senate races in this election cycle.  As the leading Senate Republican, McConnell would likely be Senate Majority Leader if the Republicans win back the Senate.  Despite his powerful position, however, McConnell appears to be running a tight race with Grimes, the current Kentucky Secretary of State.  The Kentucky race has been highlighted across numerous national news sources because of these factors, yet one rarely hears about the potential impact that either could have upon the federal judiciary.

Should Mitch McConnell win, his long-serving status could potentially aid him in nominating even more federal judges under senatorial courtesy, especially if a Republican were to win the presidency again.  Further, if the Republicans win the Senate this year and McConnell becomes Majority Leader, he would likely allow Republicans to filibuster the president’s federal court nominees.  On the other hand, if Grimes wins, President Obama would likely defer to her (as the only Democratic Senator from Kentucky) for any judicial nominations.  However, her lasting impact would be determined by whether the Republicans win the Senate.  If they do, there is no guarantee that other senators will continue to extend “senatorial courtesy” in this contentious political environment.  And should she win this election, but a Republican win the next presidential election, the president would defer to Rand Paul for nominees instead.

Unfortunately, voters rarely take the judiciary into account when deciding how to vote.  It is clear, however, that senators can have a lasting influence upon judicial policy in our country.  Thus, you may want to take into account these factors when deciding whether to vote for McConnell or Grimes in this upcoming election.


Why Should You Vote?

A recent article claims that turnout in the upcoming midterm elections this fall could be lower than the past two midterm elections, based on current voter engagement. This should be alarming considering voter turnout is already only around 40% among the voter eligible population. This predicted low level of voter turnout is based on the following three indicators of voter engagement in midterm elections: how much thought Americans have given to the elections, their expressed motivation to vote, and their enthusiasm about voting compared with past elections, which are all showing levels lower than past midterm elections. There is a positive correlation between greater voter engagement on these measures and higher voter turnout, thus predicting that fewer people will turn out to vote this November than did in 2006 or 2010 because these indicators are all lower than usual.

Why vote then? What incentive is there for Americans to turn out this November?

There are currently around 200 million Americans who are eligible to vote, only around 145 million of them are registered and of those 145 million only 131 million or so turned out to vote in the last Presidential election. We can expect significantly fewer (around 15-20%) people to turn out November 4th. People cite a number of reasons for not turning out to vote including being too busy or being uninterested. However, this is an incredibly important American right, one that everyone should be excited about since there are a number of countries that don’t have this right.

There are a number of other reasons to vote this coming election. Voting is a way to speak your mind and let your voice be heard. Voting is literally telling elected officials how you feel about important issues being debated. If that does not persuade you, know that there is power in numbers and when you vote you truly make a difference. Elections can bring to fruition incredible people; it was through elections that we voted in officials who were champions for civil rights. Our vote is our opportunity to make a different in our world.

Even though there are a number of reasons to care about elections, a huge portion of Americans STILL do not vote. So if the rest of America seems so uninterested in these midterm elections specifically then why should they matter to you?

On November 4th, Americans across the country will vote for 36 senators, 36 governors, and numerous policies. Eyes will be on the Senate races as the Republicans have a good chance (around 65%, depending on the poll that you’re looking at) of taking control from the Democrats, an issue particularly pertinent to Kentuckians who might be re-electing a Senator who could become the Senate Majority Leader.

Other than the Senate race, these midterm elections will cover marijuana legalization, abortion issues, and raising the minimum wage. Arkansas, Alaska, Nebraska, and South Dakota all have ballot initiatives that would raise the minimum wage in those states and all seem to have a good chance of passing. Alaska, Oregon, and Washington, D.C. will vote to legalize marijuana while Florida will vote on whether to legalize medical marijuana. Colorado and North Dakota both have initiatives up for vote that, if passed, would legally define life as beginning before birth and would move the respective states’ right on abortion.

Thus, even if you do not care about the turnout of the McConnell/Grimes race, it is still important to care about these elections as is true of any elections. It is a fundamental right of Americans and one that citizens should be excited about exercising.

What Has Obama Done for McConnell?

A recent article from USA Today quotes Norman Orstein, a resident scholar at the American Enterprise Institute, a Washington think tank, contending that “McConnell has become much more a national figure as a consequence of the Obama presidency than if he had spent the last several years under a John McCain presidency or a Mitt Romney presidency.” Essentially, he claims that McConnell’s tenure in the Senate can be divided into two periods: before Barack Obama became president and after. When Obama’s presidency began in 2009, McConnell had the opportunity to define himself in stark contrast to the Democratic president. Obama became McConnell’s political target against which he could contrast his party’s views.

Throughout his campaign, McConnell has advertised that a vote for Grimes is a vote for Obama. In a state where the president’s approval rating is exceedingly low at a mere 32.8%, this claim can certainly adversely affect the chances of the Democratic candidate.

In a previous post, I discussed how midterm elections are often considered a referendum on the sitting president and his or her party’s performance. According to trends, the president’s party tends to lose seats in both houses of Congress and over the past 20 midterm elections the president’s party has lost an average of 3 seats in the Senate. A continuation of this pattern is likely what McConnell is hoping to see in two weeks. In order to become majority leader, McConnell will need to win re-election but also for the GOP to see a net gain of six seats in the Senate.

However, a recent poll found that McConnell leads by the slimmest of margins, pulling ahead 44-43 among likely voters, which contrasts strongly to other polling data that shows McConnell with a four point advantage. This recent polling information contains both good and bad news for both candidates but for Grimes. Steve Voss, a Political Science professor from the University of Kentucky says, “It is clear some portion of these people who dislike President Obama, nonetheless, are showing willingness to vote for her.” In short, some people are not necessarily buying McConnell’s argument that a vote for Grimes is a vote for Obama.

Although McConnell has created an image for himself as the anti-Obama, how much this will affect Grimes depends on how convincing the argument is that voting for Grimes is like voting for Obama. It is important to note that partisanship is the primary predictor of voting patterns in U.S. Senate races and for most Democrats, an association with Obama is not the worst thing. It would seem from recent polls that show McConnell and Grimes nearly head-to-head, that the Obama/Grimes argument may not be as convincing or as influential as McConnell might hope.

Undecided Voters and Vote Allocation: Where Do Their Votes Go?

In these last two weeks before the election, the Bluegrass Poll has Senator Mitch McConnell and Secretary of State Alison Lundergan Grimes at a near even split of 44-43 in McConnell’s favor, though still within the margin of error.  As stated in previous posts where we discuss polling data, it should be noted that while this poll has the two candidates a single point apart, in the aggregate, compiled by Real Clear Politics, McConnell is at a +4. While McConnell may be ahead when looking at the aggregate polling, the question I want to explore is not which candidate polls better than the other, but rather to what end will those who have reported “undecided” make on this election. Luckily, there is research which aims at this question. With the most recent Bluegrass Poll, 8% of voters have claimed that they are “undecided” voters, and 10% of voters said that they might be willing to change their vote as the campaigning season comes to an end…so, lets see what effect this may have on the election.

One school of thought, says that the vote share of undecided voters goes against the incumbent, and the majority of the votes are cast for the challenger. This study, conducted by a private polling firm, suggests that it is the challenger who benefits from undecided voters, showing that when looking at 155 cases, 81% of the time the challenger ended up with a large allocation of votes from undecided voters. So, with that, can we chalk one up for Grimes? I would say that is not clearly the case. When looking at this study, we are not aware of what level of elections they used in their study (Local? State? National?), nor are we aware of whether these races were even competitive. In a race as close as the Bluegrass Poll is making the McConnell/Grimes race to be, I think I would like to know whether the 81% of cases where undecided voters placed a larger share of their vote for the challenger rather than the incumbent were from the election of the county jailer or for a competitive, national race. So, for Grimes, this certainly would make a great headline, but I am not convinced that this is enough evidence to support the claim entirely—we need a better study.


Luckily, we have just that. Aaron Blake of The Fix looks at the same question—where do the votes go when cast by undecided voters? Blake looks extensively at 25 competitive Senate races over the last four election seasons and sees that a larger portion of the vote share goes to the incumbent, not the challenger, when it comes to undecided voters.  He cites that only in 9 of 25 cases did challengers gain a larger vote share by undecided voters than the incumbent. Blake also shows that in his study, on Election Day, the average incumbent saw his/her vote share rise 2.5 points as compared to late polling data, and that challengers on average only saw a rise of 1.6 points. This would suggest that we would chalk one up for McConnell, but is that what will happen? While this study was communicated much more clearly than the previous, it still suggests that a good portion of the time challengers will be on the receiving end of undecided voters. Thus, I may give more support to this study and the previous, though I still would not say that the answer is conclusively that we chalk one up for McConnell when almost 40% of the time that may not be the answer (though, the fact that of the 25 studies never once did a challenger who polled behind the incumbent win the election certainly gives support for McConnell).

So, where does this leave us? It certainly leaves us on the fence about where the allocation of undecided voters will go, but there is one commonality found among both of these studies—it will more than likely not be the case that the 8% of undecided voters or 10% of those who said they would be willing to change their vote in the final weeks will split 50-50 for the candidates, as indicated by the Bluegrass Poll. Both of these studies show that there is generally an unequal divide among where undecided voters place their vote, and this could be quite an advantage in the race. Now the question we have to ask is this: will these undecided voters even show up to the polls? I guess we will find out come November 4th.

Loss of DSCC Funds May Lead to Greater Problem for Grimes than She Claims

Yesterday Kentucky was informed that the Democratic Senate Campaign Committee (DSCC) would no longer be using funds to run television ads on behalf of Secretary of State Alison Lundergan Grimes in Kentucky. This news was received at an interesting time, considering just last week was the first time since August that Grimes was polled ahead of Senator Mitch McConnell (though still within the margin of error). The question remains: why? And secondly, what effect will this have on the Grimes campaign?

One answer given by journalists as to why the DSCC will not be funding television ads on behalf of Grimes at the time is due to her answer during the KET debate Monday night where Grimes, once again, refused to say for whom she voted in 2008 and 2012. This news does not resonate well with Kentuckians, who have a deep dislike of President Obama, and perhaps the DSCC found this a good enough reason to pull out of the television wars wave between the candidates.

A second answer given is perhaps the DSCC has come to the conclusion that many others have said all along as a Democratic candidate runs against the five-term incumbent: Kentucky’s electorate is too set in its ways to move on partisanship at the national level. Chris Cilizza of the Washington post writes: “The reality of the Kentucky Senate Race is that the electorate is simply locked in, polarized to the point where persuadable voters are non-existent.” Thus, when your electorate has already made up its mind, as Cilizza has proposed, perhaps the DSCC has decided to move its efforts to races where the money poured into the race will have a greater effect.

A third answer, which ties in to the second, is that perhaps there has been a loss in confidence for Grimes’ campaign by the DSCC. This certainly would not bode well for her campaign if the DSCC came forth with that answer, and would perhaps even have a ripple effect on other large donors. The fact of the matter is that it is easy to get Grimes’ numbers in the upper 40s, but getting her over the 50% mark has proven to be much more of a difficult task in Kentucky. Perhaps the DSCC has seen just how hard this is, and thus has lost confidence in Grimes considering she has not polled over the 50% mark yet in this election, and there are fewer than three weeks to go.

While there are numerous reasons as to why the funding has been cut, the more important question is what effect this will have on Grimes’ campaign? The most obvious answer is that there will be one less group filling our television space with ads. John Sides of FiveThirtyEight writes that it is the volume of television ads which matters most, not the content. That means that it matters more about which candidate can outspend the other in television ads, so that they can fill more air space. So far, the DSCC has spent $2 million in Kentucky. While not all of that $2 million has gone to television ads, it can easily be guessed that a large amount of it probably did considering the cost of television advertisements. Putting this in to perspective, the latest released numbers on television advertising had Grimes spending $3.7 million in advertising, and McConnell at $8.1 million…thus, whatever portion of the $2 million that the DSCC was using to fund television ads on behalf of Grimes will greatly be missed as she continues to lag behind McConnell in advertising dollars.

What does Grimes think of all this? She says that “”[McConnell] can buy the airwaves, but he can’t buy the hearts and minds of Kentuckians.” While this may sound like a sufficient political response to the loss of funding, the reality is, that may just not be true. Historically, as Sides has argued, those challengers who can outspend the incumbent on a campaign by a very wide margin and those who can put up the dollars in advertising are the ones who are victorious. That is not to say that money guarantees a secure win, but it certainly improves a candidate’s chances of winning significantly, especially when there is a wide difference in spending and advertising between the two campaigns. Thus, Grimes may be in more trouble than she leads on if another source of revenue does not come to take the place of the DSCC television ads, and better yet, increase her dollars in advertising significantly to compare to McConnell.

Interpreting Grimes’ +2 Internal Poll and Bluegrass Poll Results

For the second time, Secretary of State Alison Grimes released internal polling information showing that she is two points ahead of Senator Mitch McConnell. Interestingly, in the same week, news broke out yesterday that for the first time since August, Grimes is also ahead two points in a Bluegrass Poll public poll. Is this a coincidence or not? For voters, this may be confusing, considering the last nine public polls have McConnell in the lead. Thus, in this post I want to look into two things: the significance of an internal poll, and the interpretation of the new public Bluegrass poll released yesterday.


Internal Polls

According to Nate Silver of FiveThirtyEight, public polls are generally more accurate than internal polls. Silver concludes that, generally, “when campaigns release internal polls to the public, their goal is usually not to provide the most accurate information, [but] they are most likely trying to create a favorable news narrative—and they may fiddle with these assumptions until they get the desired result.”  He goes on to say that while campaigns can spin polls just like any other organization, when numbers are released internally, it should perhaps be thought less as a scientific survey, and more of a set of talking points.

Consider, for example, a recent example of an internal poll published by Eric Cantor showing him having a 34 point lead over competitor David Brat. This is a strong example of how internal polls may not be accurate, considering Cantor was later defeated. Just as Silver says, many times internal polls should be thought of as the outlier, and not as representative as an aggregate of public polls.

So is that to say that internal polling shouldn’t be performed? Internal polling, though at times misleading to the public, does have benefits. For example, with internal polling, candidates are able to see how the public responds to the effectiveness of an advertisement. They can ask questions pertaining to the issue/ad itself, and see how separate demographics or regions respond. This is useful information to the campaign, itself. Thus, internal polling should not be dismissed, but when it is reported to the public, as said by Silver, take it as a talking point rather than a scientific survey.

Bluegrass Poll

With that preliminary knowledge on internal polls, what are voters supposed to think now that Grimes’ internal poll has her at +2 and the new Bluegrass Poll also has Grimes at +2? First and foremost, it should not be thought of as directly related Just as Silver says, there are generally different motivations when internal polls are released as opposed to public polls. Secondly, it should be noted that the Bluegrass Poll which has a Grimes lead is still within the margin of error. That is to say, there is a possibility that she is above McConnell, though it is not fully statistically supported, and therefore this could also mean that she is not in the lead and perhaps McConnell is. One of the largest misconceptions about this poll is that journalists are tweeting/emailing/facebooking that this poll shows +6 point swing from the last poll conducted in August. While the raw numbers may suggest this, when looking at the margin of error in both the polls, that statement very well may not be true. It is crucial that voters interpret the margin of error in these findings.

So what does this mean for the Grimes +2 internal poll right now and the new Bluegrass poll? It should be a place of discussion, but not a predictor of the outcome, considering McConnell is still averaging a +4.2 advantage overall in all public polls. While we cannot completely rule out that the internal survey results published by the Grimes campaign are correct, historically internal polls are unreliable as predictors of eventual election outcomes, especially when they conflict with the aggregate of other publicly released polling results. As for the Bluegrass poll, this should also be a point of discussion, but also not necessarily a predictor. Voters should have some skepticism with this poll until there are a series of other public polls predicting the same thing. As for now, McConnell has the overall advantage, and this one poll putting Grimes in the lead should not be enough to convince us that it represents a real trend in voter attitudes in Kentucky. More surveys will be needed to see whether this is a real shift in support toward Grimes or merely a statistical blip.