Are we our Parents’ Political Offspring?

August 20, 2020

I mentioned in a previous post ( how our parents shape our political beliefs. This process is part of the socialization process where children learn acceptable ways to function in society. Partisan behavior is part of this process for most children. It is a predominant factor in how adults develop partisan beliefs and loyalties.

I was reviewing a dataset from the American National Election Studies (ANES), which is a time series of surveys conducted since 1948. This file contains over 48,000 interviews with hundreds of questions on political subjects and candidates. In reviewing the data I discovered that this file contains questions about the respondent’s party id and the party id of their parents.

I realized that such a large survey could shed light on the respondents’ party id with both of their parents. If the theory of partisan socialization is correct, we would expect a high correlation of party identification between the child and parents. Below is a basic table that shows the party identification of both Father and Mother, and the corresponding party id of their child.

1. Democrat55.2%53.3%52.2%
2. Independent (some years also: shifted around)7.7%8.1%12.0%
3. Republican30.3%28.9%35.8%

As we can see, the party id percentages of Father, Mother and respondent (child) are very similar, which suggests that there is a correlation between parents’ and their offspring’s party id.

With over 17,000 interviews, the correlation between a father and the respondent is .480 (sig <.001)) and his or her mother is .439 (sig, <.001). These correlations are both strong and significant, but not perfect.

Regression analysis confirms this observation. Although the coefficients are significant at the <.001 level, the Rsq is only .258., which means that both mother and father party id only explains 26% of the variance.

This suggests that it does not completely describe why the respondent chose their party id alone. Simply put, the parents party choice does impact their child’s partisan choice significantly, but there are other influences as well.

This brief analysis does confirm the importance of our parents influence on our partisan beliefs, but other life experiences also play a role as well. It is perhaps better to look at it as a foundation that helps shapes our future experiences.

When my youngest children were pre-teens and riding in my car, my son, who was a couple of years older than his sister, said he had something very important to tell me that I might be upset about. At first I was concerned, but I assured him that whatever it was, he could tell me without worrying about how I might feel.

Hesitantly, he said, “Dad I think I’m a Republican….” At that point, his younger sister almost jumped over the front seat and shouted “See, I told you he was different from us….” Be safe…


The Margin of Error and What it Really Means

August 18, 2020

Caution: Some Math Involved

The most misunderstood statistic in politics is the “Margin of Error.” As most people know, this statistic is based on the size of the survey sample and not the size of the population. It doesn’t matter, for instance, that you are polling voters only in Fort Lauderdale or the entire United States. The sampling error for each stays the same no matter the population size.

To illustrate how to understand the margin of error, let’s say Candidate A in our poll has 48% of the vote and the margin of error for this survey is +/-3%. That means the range for Candidate A’s percent of the vote is within the margin of error, when it is between 45% and 51%. What that tells us, is that if you ran that poll 100 times, the outcome for Candidate A would be within that range, with a confidence level of 95%.

But this does not tell you if a candidate’s lead over the opponent is outside the margin of error, which would indicate his lead is greater than we would expect from sampling error. (I told you that math was involved).

If we want to know if Candidate A’s lead over Candidate B is outside the margin of error (for a two person horse-race), we have to assign the same 3-point error to candidate B as well. That means the 3-percentage margin for each candidate now becomes a +/- 6-point margin of error for the difference.

We could reasonably expect their true position to lie somewhere between –1 and +11 percentage points. In other words, for a candidate’s lead to be outside the margin of error, it must be greater than 6 points. The larger margin of error is due to the fact that if the Republican share is too high by chance, it follows that the Democratic share is likely too low, and vice versa.

In this example, for Candidate A’s lead to be outside the margin of error, it must be greater than 6 points (2 x 3%). Candidate A has 48% and Candidate B has 43%, a five point lead which is not greater than 6%, so we cannot be certain that this difference is not due to sampling error. Easy, right?

Normally you would have that explained by the reporters covering the polls, but unfortunately most don’t understand it. Many just report the single candidate margin and assume it includes the candidate’s lead as well. It doesn’t.

A lead that is inside the margin of error does not necessarily mean it is not correct. It only means we can’t be sure that it is not caused by sampling error. Polling was never meant to be a precise measuring instrument. That would require interviewing every single voter. It is an estimate within certain parameters based on probability theory.

Occasionally polls predict the exact the outcome. When that happens, pollsters pat themselves on the back and tout their great survey techniques, when in most cases it’s just dumb luck (random). I know, I’ve done it… Be safe.


Most Voters think that we have two Parties with some Independents. They are wrong!

August 16, 2020

I have been negligent not to cover this subject before because it is one the more important tenets of modern political science. The concept of party identification is one of the most important theories in American politics and still the major reason for most people’s vote choice.

But most Americans think that there are only two parties, with some independents stuck in between them. Admittedly, this a simple concept and understandable, but it is wrong.

Political Scientists have understood since the 1960’s with the development of modern polling, that partisan identification is a continuum of party identities and strengths and not a binary choice.

One the biggest misunderstanding is that independents are a homogeneous group that makes their choice absent of partisan influences. But studies show that this not the case.

From the 1960’s on, political scientists have developed methods to uncover the multiple nuances of party identification. The longest and most commonly used method is known as the Michigan scale.

With this scale, Party identification is measured by asking individuals whether they consider themselves to be a Democrat, Republican, or independent. Those indicating Democratic or Republican are then asked whether they are a strong or a weak Democrat or Republican, while those claiming to be an independent are asked whether they feel closer to one of the two political parties.

This yields a seven-fold classification: strong Democrats, weak Democrats, independents closer to the Democrats, independents not closer to either party, independents closer to the Republicans, weak Republicans, and strong Republicans.

Using data from the 2020 American National Election Survey (ANES), I can test the impact of this party id scale on candidate impressions and vote choice.

Using their thermometer ratings on Trump and Biden, where voters rate on a scale of zero to 100 degrees (positive ratings above 50 degrees), we can see how these seven different party identities rate these two candidates for President. Using the seven-point party identification, we find significant rating differences between Republicans when we separate them into different partisan strengths.

The strong Republicans rate Donald Trump with a remarkable 85 degrees. If he was arrested tomorrow, they would still vote for him. But as levels of Republican strength decline, so does his rating. There is a drop of over 20 degrees between the strong and weak Republicans. But this is still a healthy rating.

With Independents who lean Republican, his rating is, oddly, slightly higher. Remember, these are voters who initially said they were Independents. But as you can see, they act like traditional Republicans.

Pure independents drop below 50 degrees, into negative territory, but still higher than independents who lean Democratic, who put the President at the second lowest rating at 17.4 degrees.

As expected, both weak and strong Democrats have negative views of him, with strong Democrats bringing up the rear at less than 15 degrees.

Before we decide how these voters will ultimately vote, we have to see how well Joe Biden does with the categories of voters. Below is a chart of the Biden ratings using the same seven point scale.

As expected, we see the opposite of Trump ratings for Biden. What stood out me is how pure independents dislike both candidates. Trump’s independent rating is 33.4 degrees and Biden’s is 34.8 degrees. For them, this is a hold-your-nose choice.

The other major difference is the different levels of intensities between Trump and Biden ratings among their own partisans. The most important difference are among strong partisans, where strong Republican’s give Trump an 85.2 degree rating and strong Democrat’s rate Biden at 70.7 degrees, a 14.5 degree difference in Trump’s favor. The same is true, but at a lesser level, of the differences among weaker Democrats and Republicans.

Intensity of support for Trump has always been his strong point, and so far, that is still the case. How would that affect the election? In a close election, it could be deciding. Bad weather, complacency, and over-confidence could effect the Democratic vote. My opinion of Trump voters is that they would show up in a Blizzard in November for him.

Political studies show a high correlation between the seven-point identification and vote choice. In this survey, they included who they would vote for in the general election if it were between Donald Trump and Joe Biden.

This survey was completed in July, so I wouldn’t put too much faith in the trial ballot, where Biden has 51.3% and Trump 48.7%. In the table below, I compare vote choice by these voters’ 7-point party id.


As you can see, Trump’s percent of the vote starting with strong Republicans shows they choose him over Biden with 96.6% of their vote. Weak Republicans drop down to 84.3% and independents who lean Republicans at 87.2%.

In other words, voters who said they were independents initially, but when probed said they leaned Republican, chose Trump almost 3% more than traditional Republicans.

The same is true for Independents who said they leaned Democratic. They gave Biden 88% of their vote compared to traditional Democrats 85.6%. Both Republican and Democratic leaner’s are more partisan that traditional rank and file partisans.

These results are confirmed by previous studies showing that many so-called independents are more partisan than weaker party voters.

The other take away is that Republican voters give more support to their candidates than Democrats. And pure Independents favor Democrats slightly more than Republicans.

I want to emphasize that these results don’t predict who will be victorious in November. But the thermometer ratings at this point in time, suggest a closer race than current polls now indicate.

This analysis confirms what political scientists have known for sometime. First, a person’s party id is the primary cause for vote choice even in presidential campaigns where voters are far more informed on the candidate’s positions on important issues than in state or local elections. In low information elections, party id is often the only reason.

The long held belief that independents are neutral is basically a myth. As I show, some types of independents are more partisan than many Democrats and Republicans.

The only problem for poll watchers is that practically no public pollsters use this scale. It is almost exclusively an academic survey question.

That’s unfortunate, because in combination with the trial ballot question, it reveals more information about a voter’s likely choice than any other traditional question. Do a political junkie, like myself a favor, call your Congressman and demand that public pollsters insert this in every political questionnaire! Be safe…


Through the Eyes of Super Trump Voters

August 12, 2020

Recently, I received the latest survey data set from the American National Election Studies (ANES), a consortium of Stanford University and The University of Michigan. It is a national online survey of 3080 adults (but contains Party designation) completed in July. As typical of academic surveys, it is a massive file containing hundreds of questions.

This particular study is called an Exploratory Testing Survey, which means that most of the questions are unique and provided by academic members of the ANES.

It provides an opportunity to look at how voters feel about issues not usually included in polls. My interest is how Trump voters actually feel about Donald Trump and his particular style of governing. In other words, questions that go beyond the usual job approval ratings.

I used the headline “Super Trump Voters,” to describe voters who would still vote for him even if he shot Mike Pence on Fifth Avenue. Like almost all academic political surveys, instead of a favorable rating, they use the Thermometer rating which ranges from 0 degrees to 100 degrees, where 0 to 49 degrees are in the cold range and those 51 and above are warmer.

I personally use the this scale in most of my surveys because it not only tells how voters like or dislike a candidate, but also the intensity of their feelings. I will admit that it is rare to see a 100 degree rating for a politician.

But in the case of Trump, that wasn’t the case. Some 345 voters gave him a 100 degree rating or almost 12% of the sample. My assumption is that any voter who rates Donald Trump a 100, is more than just a Trump voter but a “super” Trump voter. It does not represent all people who support Donald Trump, but only those who see him like a famous rock star.

The survey included questions on what I would call character assessments, such as leadership and that he cares about people. In the first table below, we see if Trump voters’ see him as empathizing with them.

Really cares about people like you. - How well do each of the following traits describe Donald Trump...					
 		             Frequency	Percent             	
	1. Extremely well	263	76.2		
	2. Very well	53	15.4	15.4	
	3. Moderately well	19	5.5		
	4. Slightly well	4	1.2		
	5. Not well at all	6	1.7		
	Total	               345	100.0		

With Trump supporters, 91.6% say the statement fits the President either extremely well or very well. This question was designed to measure empathy and most politicians would kill for this rating.

Table 2 Dignified. – How well do each of the following traits describe Donald Trump?
1.Extremely well20860.3
2. Very well8424.3
3. Moderately well3911.3
4. Slightly well82.3
5. Not well at all61.7

We expect our President, of course, to be dignified and it relates to his ability represent the United States to the world. It is a trait that his favorite President Andrew Jackson did not have. His inauguration party at the White House attracted some 20,000 supporters who drank every bottle of liquor they could find and it took three weeks to clean the house up.

Table 3 Honest. – How well do each of the following traits describe Donald Trump?
1. Extremely well24370.4 
2. Very well6819.7 
3. Moderately well247.0 
4. Slightly well72.0 
5. Not well at all3.9 

And, of course, we expect our President to be honest, just like Honest Abe Lincoln, who’s likeness is carved into Mount Rushmore and where there is still room for another president.

In Table 4 below, some 82% say the statement he is “Authentic” fits him extremely extremely well.

Table 4 Authentic. – How well do each of the following traits describe Donald Trump?
1. Extremely well28382.0  
2. Very well4011.6  
3. Moderately well113.2  
4. Slightly well51.4  
5. Not well at all51.4  
99. Missing1.3  

I know some might disagree with this statement, but I would have to say that having spent half a day with him when he was renovating Mar-A-Largo, he maybe not authentic but he was certainly different.

Table 5 Divisive. – How well do each of the following traits describe Donald Trump?
1. Extremely well14441.7  
2. Very well5816.8  
3. Moderately well329.3  
4. Slightly well267.5  
5. Not well at all8524.6  

I was somewhat surprised that as many as 42% Trump supporters called him “divisive.” Divisive is not a superlative people normally use for someone they respect or admire. That said, it is in my opinion it is accurate and indicates that his base likes this aspect of him.

Table 6 Knowledgeable. – How well do each of the following traits describe Donald Trump?
1. Extremely well26576.8  
2. Very well5816.8  
3. Moderately well144.1  
4. Slightly well51.4  
5. Not well at all3.9  

Here is where I disagree with his avid supporters’ opinion. Donald Trump is not a reader and history is not his strong point. In fact, he has made fact checking a full time profession.

Very few studies so far have looked into the opinions of Trump’s loyalist base. The current social psychological evaluations suggest that some Trump supporters lean toward authoritarianism. But if nothing else, this data paints a picture of Americans who see a decisive leader that can on occasion be divisive.

I have to admit that my short time with him in the late 1990’s, revealed a man dead certain about his opinions. I could tell I was irritating him by suggesting that more study of the subject was needed. As I was about to leave, he stood up and pointed his finger at me and said “if I do it they will come…” When he ran for President he was right. They came and voted for him. Be safe…


The Political Polarization of Abortion

August 7, 2020

It wasn’t that long ago when Republicans and Democrats generally agreed on when a woman could obtain an abortion. Like most issues in the U.S., politics made us reconsider the issue in a partisan way.

In the 1972 presidential election, Richard Nixon saw a potential way to attract Catholic votes. At that point in our history, Catholics were the most anti-abortion group. So he began a campaign strategy to take anti-abortion positions to attract them and also social conservatives.

In the 1972, Nixon won the election and carried a majority of Catholics. With that success, Republican strategists began to apply this strategy to Congressional races, using coalitions of evangelic voters around the abortion issue.

At this point, both Republican and Democratic rank and file voters we in near agreement on when an abortion should be legal. As an example, from 1977 though 1985, with the exception of 1978 and 1985 were they tied, more Republicans supported an “abortion for any reason” than Democrats, as shown below.


But the tide turned at the end of Ronald Reagan’s second term. In the 1980 election campaign, Reagan made it a campaign promise to appoint anti-abortion judges. His position on abortion slowly seeped down into the rank and file beliefs, as shown in the table below, the percent of voters saying yes that a “woman should be able to have an abortion for any reason” declined among Republicans and increased with Democrats.


As you can see, the big change occurred in 1988, where Democrats overtook the Republicans in their support of woman’s right to have an abortion for any reason.

In the chart below, using the data from our table, you will see how the division between Democrats and Republican changed in an exponential way.

Blue line = Democrats / Red line = Republicans

In 1990, partisan voters went their separate ways. The Republican support has leveled off around 29%, while the Democrats have continued to increase their support for a woman’s right to have an abortion, with over 60% now supporting this view.

Are woman more pro-choice than men? You might expect a woman to be more pro-choice than a man but that assumption would be wrong. Using the most liberal of reasons for an abortion, that a woman should be able to decide an abortion for any reason, it turns out that both men and woman agree on this issue, as shown in the chart below.


The two lines nearly strangle each other and the average difference is only 1.3%. Obviously, these men are married and know when to agree.

Women and Men Percent Yes on Abortion for any Reason

1977 – 1918

women % Yesmen % Yes

Donald Trump has carried on the Republican tradition of appealing to pro-life voters by announcing during the 2016 campaign that he would appoint only judges who would overturn Roe v. Wade. And so we are now back to the basic strategy started by Richard Nixon in 1972. Somethings never change…Be safe.


Partisanship and Voting Choice: The Myth of the Reasoning Voter

August 5, 2020

If there was one book I would recommend for a serious student of political behavior, it would be the The American Voter. ( A., Converse, P. E., Miller, W. E., & Stokes, D. E. (1960). Published in 1960, it laid the foundation for all studies on political behavior since.

The cornerstone of their analysis was the concept of partisan identification. Their understanding of party identification goes beyond group attachment like a membership in a club or group, but a psychological identification, “which can persist without legal recognition or evidence of formal membership and even a consistent record of party support.”

In other words, partisanship is a social identification such as religion or an ethnic group, which when invoked can sometimes cause emotional reactions. My mother told me when I was a young to “never to discuss religion or politics” with friends. And she was right.

Most scholars agree that party identification is formed by socialization in your youth, primarily from your parents. There are some recent studies testing the genetic aspects of political behavior, with experiments with identical twins. Although the results are interesting, it does not yet show a significant correlation but likely will in the future.

In an attempt to demonstrate the impact of partisanship on candidate choice, I have collected voting data from 13 presidential elections, 1968 through 2016. Below is a chart of how many voters chose the Republican candidate by their party or non-party affiliation.

Presidential vote for Republican candidate by party identification. Data from General Social Studies survey.

The top line (grey), represents Republican voters percentage for the Republican candidate. For example in 1976, 84.4% of Republicans voted for their party’s choice. The Democrats (darker blue, bottom line) in that same election, only 15.1% chose the Republican candidate. The middle line (light blue), shows how many independent voters selected the Republican.

The average Republican vote for the Republican candidate over these 13 elections was 87.8%. The average Democratic vote for the Republican candidate was 14.5%, an average difference of 73.3%.

There a couple of interesting deviations in this chart. In 1992 and 1996, all three group’s percentages for the Republican candidate declined significantly. In 1992, Bill Clinton defeated George H. W. Bush by almost 6% in a three was race with Ross Perot.

Bush’s last approval rating was 32% just prior to the election. It’s actually amazing that 75% of Republicans voted for him and it illustrates the power partisanship, even though some obviously defected from their standard bearer.

The Republican vote rebounded in 2000, with 94.2% voting for George W. Bush. After that they stayed the high range with an 88% vote for Donald Trump.

Independents also took a dive from Republicans in 1992 and 1996, but rebounded in 2000, splitting their vote for Bush and Gore. Over the 13 elections, they averaged 43.1% of the vote, which means they voted for Democrats by nearly 47%. That would imply that they lean slightly toward Democratic candidates.

The effect of partisanship is most apparent from polls of Trump’s Job Approval ratings. In the latest national survey by Civiqs of registered voters, shows that Trump’s current job approval rating is 41% (56% disapprove), which is consistent with other major polls.

Missing values “don’t know”

If we break down that rating by party identification, it reveals that his approval rating from Republicans is now 87% (9% disapproval). For Democrats it is 3%, a difference of 84% ! That is a real partisan divide.

This partisan divide suggests that Trump’s path to victory will have to come from Republican and Independent voters and forget about winning enough Democrats to swing his way. It is likely that he would need 90% or more of the Republican vote to have any chance of winning the National vote.

The author’s of the American Voter demonstrated that most voters had likely decided on which candidate to vote for in 2020, when he or she graduated from High School ! They just didn’t know about it yet. With the possible exception of independents, the idea of a free thinking American voter has always been largely a myth. Be safe…


Have the Democrats and Republicans lost their Convention Bump?

August 1, 2020

When I was growing up, the Party Conventions decided the nominee through a series of votes by delegates who were wheeling and dealing for influence. It was a pageant that would put any of today’s reality shows to shame.

I remember watching it with my father on a black and white TV, staying up longer than my mother would normally allow. It was both a spectacle and an education for a young boy watching delegates cutting deals with each other on live television. I loved it.

But that was a long time ago, and modern conventions now provide a platform to introduce their nominee, already decided by party-primary voters, before a national TV audience covered in a rally like demonstration of party unity, balloons and all.

But the conventions still provide a significant function by writing platforms, introducing the vice-presidential nominee and showing party unity after a sometimes contentious primary campaign. If successful, the warring candidates and their supporters, end up hugging each other and that maybe it’s most important contribution to the party’s nominee’s success, as Clinton’s campaign discovered in their 2016 Convention.

But what maybe the most important aspect of a successful convention is what academics and political observers call the “the Convention bounce.” This is the increase in the nominee’s polling numbers following the convention.

Previous studies on a convention’s impact before 1988, have shown that the difference between trial heat polls prior to the convention and after are typically around 6%. (The Convention Bump, American Politics Quarterly, July 1992).

But Covid-19 has changed the rules for this year’s conventions. No longer will a wild and enthusiastic crowd waving signs, festooned with descending balloons broadcast on national TV, and interrupted occasionally by celebrities and party leaders.

Instead, the plan for the Democrats is to conduct a “virtual convention,” which means the state delegates will be voting from their homes for the already decided nominee, Joe Biden. No pageantry and no balloons. Joe Biden will make an acceptance speech in Milwaukee, but not in front of screaming delegates acting like teenage girls as if Joe was Justin Bieber. (I will miss this part.)

Donald Trump has given up on his Party’s convention in Jacksonville due the out of control virus in Florida. He says he will accept the nomination back in Raleigh, North Carolina, after pulling out of that state because of pandemic restrictions (masks?). The details are still, at this point, a little murky but I’m sure it will be broadcast on national television.

But these changes provide an opportunity to see if the pageantry of the previous party conventions really make a difference in the degree of a ” convention bump.”

But first we need to see how previous conventions effected their party’s nominee in the trial heat polls. Below is a table of how previous conventions effected the presidential candidates from 1980 through 2016.

198025   REP
198436   REP
198877   REP
1992170  DEM
199458  DEM
199685  REP
200004  REP
200852  DEM
201234  DEM
201622  REP
Red = Republican Candidate / Blue = Democratic Candidate

This table and chart shows the percent change in the party’s nominee’s trial heat polls in the Gallup survey two weeks prior to the convention as compared to the first Gallup poll after the convention. For example, in 1980, the Democratic candidate, Jimmy Carter, gained 2% in the post convention poll (a 2% bump). Ronald Reagan had a gain of 5% between his prior convention poll and the post convention poll (a 5% bump).

In 1992, Bill Clinton made his first campaign for president against incumbent President H. W. Bush. In the month before the convention, his polling share of the averaged 25%.

He was practically an unknown until the convention, but he was charismatic and he could “feel your pain.” When the dust settled after that convention his percent of the two party vote went from 40% to 57%!

The final column lists the party winner in each campaign. In our 1980 example, the winner indicates the Republican party nominee the winner (Ronald Reagan).

The average convention bump for all 10 election cycles for Democratic candidates was 5.2% and for Republican candidates 4.3%. No candidate in this time period had a negative result and two candidates had a net gain of zero.

Out of the ten election conventions, the candidate with the largest poll increase won six elections, suggesting that a positive convention develops some momentum.

In 2016, Hilary Clinton and Donald Trump had identical convention bumps of 2%, which may have foretold a close election.

Now that we have examples of previous “convention bumps” we can measure the changes caused by an abbreviated convention format.

If after the convention the bump is still within the average of previous election cycles, we can conclude that the changes had no effect on the election bump. If, however, it is significantly lower than previous conventions, we can conclude that having a more traditional convention is politically advantageous.

We will have to wait for the two conventions before we can calculate the impact. I don’t know about you, but I’m betting on a convention dud. Be safe…



July 29, 2020

George W. Bush’s first term started with an approval rating of 60% and after the terrorist attacks on 9-11, it reached as high as 92%. But from that time on, his approval rating slowly deteriorated until it reached a low in May of 2004 at 46%, just five months before his reelection against John Kerry.

Both Bush and his political consultants at the time, understood the importance of the approval rating. No incumbent president had won reelection with a rating less than 48% since Truman. Bush’s father was defeated in 1992 with a 34% rating and now he faced defeat himself.

As the table below shows, out of the past 10 presidential reelections, only three were defeated and all three had ratings less that 48%.

ColumnYearJune of reelection yearFinal measure before electionWon reelection
  % Approve% Approve 
G.W. Bush20044948Yes
G.H.W. Bush19923734No

But George W. Bush’s approval rating reached the magical number just prior to election day, and he defeated Democrat John Kerry by a whopping 2.4% and 286 electoral votes.

It’s now the end of July and the latest national polls put Trump down by an average 9% and a current job approval rating at 42% (Gallup has him at 38%) with only a little more than three months till election day. The chart below shows both Bush and Trump’s approval ratings during the same time period.

Bush rating blue line and Trump rating red line.

Only Obama was able to increase his rating by 4% between June and November. Even if Trump manages that increase, it still only gives him a rating of 46%.

So forget about the trial heat polls, just look for the last job approval rating before the election. If Trump hit’s 48%, call your bookie and put it all on Donald J. Trump. Be safe…


Another Metric that can Predict the Presidential Election Winner?

July 26, 2020

When I was doing research on how bookies make odds on elections, I found an article that noted how some bookies use favorability ratings of presidential candidates to calculate their odds. I personally have used these ratings in local and state surveys and found them to be quite useful at that level.

So I decided to see if these ratings had any impact on presidential races in the past. From Monmouth University’s monthly blog, I found data from their surveys going back to 1984, that included favorability ratings.

Their question gives voters three choices: favorable or unfavorable for Democratic candidate, unfavorable or unfavorable for Republican candidate, or neither candidate favorable. This question is strictly a popularity measure, but likely contains some perceptions of job performance.

The table below shows how voters ranked Democratic and Republican presidential candidates from 1984 to 2016 (this question wasn’t asked in 2008).


What jumps out at is the 2016 election, where 35% chose neither! Monmouth says they have not seen this kind of reaction since using this question. But this was an unusual pair. Both Trump and Clinton had the lowest favorable ratings of all presidential candidates since 1984.

But is there a relationship between favorable ratings and winning the election? Monmouth did not address this question but my curiosity forced me find out.

To test if candidate ratings effect the election, I have added a column for the winning popular vote candidate, where 1 represents the Republican candidate and 0 the Democratic candidate. It does not measure the Electoral College vote.

In all of these eight elections, the popular vote was won by the candidate with the more favorable ratings, even if it was as little as 1%.

The average of Clinton’s favorable ratings is 38.9% and Trump at 37.4%, a difference of only 1.5%. This difference isn’t great, and a test of the mean differences suggests that the means are are equal to or greater than each other (p=<.001).

So how do Trump and Biden favorable ratings compare at this point in the campaign? Using Real Clear Politics average of recent favorable ratings (July/June), shows Trump with an average ratings of 39.5% and Joe Biden with an average rating of 45.5%, a difference of 6%.

Like most polls, Job Approval ratings, betting odds, Job Approval ratings, and now this Favorability rating, all give Biden, at this point in time, a significant chance of winning at the national level. This does not include the swing states and the Electoral College.

Remember, we still have over three months until the finish line. Polls are most accurate when nearer election day. Go ahead place your bets, but don’t bet your house on it just yet.

In a later post, I’ll explore the favorability ratings in the swing states, which if Trump is going to pull out a victory it will likely occur there by winning the Electoral College vote. Be safe…


Do Likely Voter Polls Benefit Donald Trump?

July 22, 2020

The trend among most pollsters has always been to use a sample of all residents or registered voters early on in the election campaign and switch to likely voters as the election nears.

The reason for this protocol was explained to me by an early leader in modern campaign polling, Bill Hamilton, when we both taught at the University of Florida.

He explained that voters need some time to absorb the effects of the campaign and the competing candidates. Using a likely vote screen early on, leads many voters to fib about their likelihood of voting in an election still months away. And besides, he would say. “it’s cheaper.”

As we get closer to the November election, almost all polling firms will be utilizing likely voter screens.

But in the polling and campaign business, there are some who believe that the method of interviewing registered voters rather than likely voters early on, is an intentional method to under-report Trump’s real support in the polls.

This argument suggests that the more liberal public polling firms, by calling only registered voters is an intentional effort to increase Biden’s poll lead.

John McLaughlin, well known Republican pollster, commented how many pollsters are skewing the results by using registered voter samples.

The refusal to screen for actual likely voters is creating an under-polling of Republicans and therefore Trump voters. It seems intentional. It’s exactly what the media did in 2016. Let’s prove them wrong again.” ( ElectionsPolls & SurveysPress Releases, June 8, 2020)

For me, this argument begs the question, what are the polling differences between a registered voter sample and a so-called likely voter one.

The idea that Republican voters are more likely to vote in elections and Democrats less likely, has been asserted for many years. The argument goes like this, Republicans are better educated, wealthier and older. These are attributes that contribute to voting.

In 2010, the website FiveThityEight concluded that registered voters surveys favored Democrats by 5% over the actual election results and likely polls favored Republican candidates by 1%.

Oddly, the academic studies have not focused directly on this issue. It is true, that the opinion of pollsters and campaign consultants are all in agreement that more likely voters identify with the Republican Party and, of course, likely to choose a Republican candidate over a Democrat.

To address this issue on whether likely voter polls do favor Trump and registered voters Biden, I devised a way to test these assertions statistically by comparing two series of recent national surveys conducted by public polling firms.

The first group includes 22 polls of registered voters only. The second group includes 22 likely voter polls only. Both survey groups were completed in June and July with similar end dates. The polls asked the identical question of choice for president.

When comparing Trump’s percent only in both registered and likely voter surveys, the results showed a small difference between the two groups. The chart shows how likely voters tended to support Trump more than the registered voter surveys.

Blue line = Likely Voter surveys /Red line = Registered Voter surveys

As you can see the per survey differences weren’t substantial, with a mean difference of only 1.73%, in Trump’s favor.

mean 41.5mean 39.7

But that’s only half of the story. This table and graph only shows the changes in Trump’s percentage without Biden. When you look at the percent difference between Biden and Trump in each poll, it reveals a completely different picture. After all, Trump isn’t running against himself (no comments, please).

The best way to measure the differences between two candidates is subtracting one percentage from another, creating a net difference variable. In this example, I’ve subtracted Trump’s results from Biden’s. The table below shows the net percentage difference between Biden and Trump by whether the sample was of registered (RV) or likely voters (LV). Positive numbers indicate a net percentage gain for Biden.

RV % DifferenceLV % Difference
RV mean = 9.2LV mean = 7.6
Biden / Trump Net % Difference in Registered Voters and Likely Voters

All the differences are positive except one likely voter survey with a minus 4%, which means that Trump had 4% lead over Biden in that one survey only. As the means for the two different modes show, Biden’s lead among registered voters is 9.2% and a 7.6% for the likely voter poll. In other words, Trump performs better in the likely voter sample by an average difference of 1.6%.

But is this modest difference real or just a random variation. To determine whether the differences are not random, I used a one sample T-test, which compares the mean of your sample data to a known value. In this case, is the mean of registered voters versus likely voters?

Without boring most of you with statistical data, the answer is yes, and the differences are statistically different at a P value of <.0001.

The individual differences between individual polls can sometimes be substantially larger than others. By comparing the results of two separate sets of polls, the differences are averaged out.

Notice that in this example, there are some polls showing Biden leading by 12% or even 14%. But when you combine those outliers with other polls, we average out those differences. (Remember the wisdom of the crowd.)

An example of that is an ABC News / Washington post poll released last Sunday. To their credit they used both registered and likely voters in the same survey. The registered voters had Biden at 55% and Trump at 40%, a 15 point differences. In the likely voter sample, Biden had 54% and Trump 44%, a 10% difference. The pattern is the same, Trump does better when the poll uses a likely voter sample, but the difference is based on only one survey.

This analysis confirms that Trump on average, does do better in likely voter polls than with traditional registered voter samples. It’s not at the level of the recent ABC / Washington Post Poll showing a 10% difference. This is not the case when we average a series of public polls in the same time period, using identical sampling methods.

With most recent public polls showing substantial leads for Biden, an average difference of 1.6% won’t help much. With almost four months and two Presidential debates to go, this campaign is just getting started. Stay tuned and stay safe…