Campaigns and Elections Final
Valence Issues
- A reason voters might vote for a candidate without the same PID - Candidates want a higher level of valence - "Slick, Wonk, Fool"
Campbell's "American Grace: How Religion Divided and United Us"
- Religion is a lot more of a divider than a uniter - We are devout, tolerant, and diverse - The US has a high rate of weekly attendance at religious services (about 37%) - Americans are more religious than Iranaians - Religious identity on the decline - 1970s to 90s is growth in evangelical churches - 90s is non religious - The nones are a direct reaction to the intermingling of religion and politics in the US - Religiosity is associated with Republicans more - Americans are generally kind about religion - Americans like Muslims the least, Jews the most - Most choose Aunt Susan to go to heaven
Hopkins' "Political Twitter Is No Place For Moderates,"
- Russian orgin division - Americans that tweet about politics come from the extremes on the view spectrum - Very conservative or very liberal is who makes up these people - The political words were reserved for the small radicals - The graph looks like a C shape - Conservatives: pjnet, Q, fox, trump - Liberals: gay, racism, Sanders (most common terms) - Only about 12% of adults use social media via Twitter - Political Tweeters are WAY More polarized than the average adult.
A big ask
- Someone asked me! The most common response from volunteers about why they're doing it.
Castle, Burge, and Djupe's "Younger evangelicals look an awful lot like older evangelicals, post -2016"
- Surveys show young evangelics to be much more conservative than others of their generation - Evangelicas 20% of population, 70 to 80 percent Republican. Very important voting bloc - NY Times article covered liberal evangelicals - Racial minorities are living little effect on aggregate partisanship - They support Trump big league regardless of age - The other ones voted libertarian, not democrat - The data that they are shifting is simply not true and doesn't show in the actual voting data. - "the more things change, the more they stay the same"
Today political media is both
- Top down (old media) - Bottom up (social media)
Semi-closed
: People registered with party and independents can vote
Megachurches
Are increasing, and their parishioners vary across many demographics
Percent of Americans in competitive districts
Has been on the decline
Hispanic Voters
Mix of Voting, but lean Dem
Between education and party ID vote choice is
Nonlinear - it's not increasing or decreasing
Unmoved Mover
Party ID is so powerful that it is known as this. It affects everything else, and nothing really shifts it.
Marriage Gap is
Somewhat due to income, married households have higher incomes are unlikely to be in poverty - GOP has law and order reputation and family values
Most important predictor according to Ferguson
Whether out party candidate is more ideologically extreme than in-party candidate
Do Campaigns Matter?
Yes, but most campaign effects are systematic or predictable
Open
any registered voter can vote
Caucuses
are run by a party's state organization • You show up with a bunch of other people and mill around for a while • Conducting party business like selecting local party leader and deciding the state party's platform • In addition to voting for a candidate
Forecasting Models
attempt to predict election results • In advance of the actual election using known quantities measurable over time • Y=mx+B • Y= incumbent party share of 2 party vote
Sophomore Surge
average gain in vote share for MCs running as incumbents for the 1st time, compared to their share in the first election
It's comforting
to believe that the other team is just duped/misled GOP, Obama winning 2012 Dems, Trump winning 2016
The most important sentence in Get Out the Vote
"The more personal the interaction between campaign and potential voter, the more it raises a person's chances of voting" (9) In today's digital world: - human interaction is more scarce but valuable - Old fashioned (pre-TV, pre-phone) campaign techniques are more valuable again - Election Day parties/parades/CARAVANAS, neighborhood meetings, events, rallies, etc are valuable
Establishment
"Top politicians and party officers... traditional interest groups, issue advocacy groups and ideological activists" EVERYONE is included in this definition of establishment
Party establishment control
"rests on their ability to reach early agreement on whom to support and to exploit two kinds of advantages- control of campaign resources (money, knowledge, labor) and the persuasive power of a united front of inside players."
Abramowitz's "Why Democrats Should Ignore Swing Voters and Focus on Voter Registration and Mobilization"
- 2012 was going to be a close one, not. - Democrats have little chance of winning over many swing voters but a better chance of winning the votes of the unregistered - Swing voters unhappy with Obama, don't vote - A small group of swing voters can easily decide the outcome - White and female are swing voters - Swing voters held very similar opinions to that of Romney supporters but were not enthusiastic about voting - Twice as many unregistered voters then swing voters - These findings suggest that the Obama campaign would be well advised to focus its efforts in the battleground states on voter registration and turnout rather than on trying to win over swing voters. However, unregistered voters, like swing voters, were rather unenthusiastic about voting. Getting them registered and to the polls could be challenging. - Obama unregistered much more excited than Romney unregistered - Turnout helped Scott Walker win in WI - Obama solid to win WI and that he did
Enten's "Voters Used to See Moderation in Trump. Not Anymore."
- A majority of voters think Trump is conservative - Trump's views were hard to pin down before he took office - Blast Trump paragraph - Trump +60 "Hard core conservative"
Combining V=(P*B)-C+D and Get Out the Vote gives you
- A way to win campaigns via turnout! - Which can happen. - Turnout was not the reason that Trump won in 2016. - Lower turnout by black Americans hurt Clinton.
Increased media choice & correlation between media choice and politics
- Allows campaigns to reach particular segments of voters ------------- even within each party
Trende's "How Trump Won - Conclusions"
- Almost all election analysis have focused on rACE - Demographic analysis is way over used - Demographic impact is often overstated. - Demographic changes are meaningless without a geographic filter - Our nation's politics at every level have a geographic filter - Trende uses a demographic urbanicity model that easily filters through a geographic component - Democrats are far too clustered in urban centers to be effective, even when they win the popular vote - Rural and small town america cast tens of millions of votes, and this adds up - California votes are wasted - CBSA divisions (geographic omponent) - Clinton did well in mega cities, and large cities - She did horrible in small town and rural America - She lost Florida, Georgia, and Pennsylvania - Rural and small town America going away is much slower than people seem to realize - 1. This process didn't happen overnight - 2. This hurts Dems in the Electoral College, as well as Senate and House - 3. This trend may be reversible
McLaughlin's "The Bump Stock Debate Illustrates Two Competing Models of Political Strategy"
- Automatic weapons are very rarely used for crime in the USA - Bump stocks simulate a fully automatic rate of fire on a semiauto gun - High ground strategy: Pick your battles or have them picked for you. - Not one inch strategy: politics is about mobilization rather than persuasion - Morale: by forcing the other side to fight to the death for everything, no matter how modest, you wear them down, drain their energy and resources, and keep your own people engaged and feeling like winners. - You give a Democrat an inch and they take several hundred miles - Working with Democrats on gun control is a very dangerous game because we know their end game. - the Pro gun side has employed the not one inch strategy and been very successful - McLaughlin continues to see the high ground as a more successful strategy. Win majorities and convert over time. It's worked on guns but that's about it. High ground, or trenches?
The internet has changed the way campaigns ask supporters to participate:
- Candidates don't need established fundraising networks as much - This gives tech-savvy outsider candidates more power at the expense of party elites - Increase web-prescence, Increase small dollar funraising, and increases electoral support + volunteering
What effects the P term
- Competitiveness of the candidates affects the P - Size of the potential electorate matters - Americans do think strategically about election closeness
Forecasting and Congressional Elections
- Congressional elections are nationalized today - Favorability of incumbent - Voters often have something good to say about their MC - When they do, they are likely to vote for him or her
What causes the gender gap in PID and voting?
- Differences in policy views - Women more liberal on: Education, health policy, social welfare, defense - Abortion is not the cause
Ladds "Two Approaches to Predicting Which Party will Win in 2016"
- Difficult to tell who GOP nom would be, Hillary was already the Dem nom - Effect of the candidate is small anyways National conditions perspective: - Economic growth in the first three quarters of the election year, number of terms the incumbent has held, and # of casualties in overseas wars - This is called predicting using the fundamentals - This model treats the parties as interchangeable The Strength of the Party Coalitins Perspective: - More identified as Democrats during the entire period but the Republicans did way better - Republicans must bridge demographic gaps to continue winning Evaluating the Perspectives: - National Conditions Perspective is better - Correctly said that Democrats victory is not guaranteed
The American Voter a(1960) argument
- Elections don't work because: - Voters don't have ideology or issues prefs - Elections don't represent a desire for policy change - Professor View v Pollster View (smart people say stupid things), peoples preferences do in fact change and they do desire policy change.
V=(P*B)-C-D
- Explains why many people vote, and why many people do not vote - Campaigns manipulate different parts of this equation to boost or lower voter turnout - The P and B terms change from election to election!
Cohn's "How The Obama Coalition Crumbled, Leaving An Opening For Trump,"
- FBI director sent a letter to Congress about Hillary (Comey) - Trump's strength was clear before Comey and Wikileaks - Iowa and Industrial belt liked Obama and then got played by the Democrats - White Voters in the Northern United States - Trump dominates white working class voters - Democratic strength eroding in states like PA, and Wisconsin - Turnout among GOP and Dems was even at 2.5 increase - Trump's views on immigration, trade, China. guns and Islam sold the package well to the WWCV (White working class voter) - Black turnout is on the decline - Young black voters are registering less and less - 5 to 10 percent drop among black voters accoss the US - Dems winning more white "Educated" college voters
Hopkin's "All Politics Is National Because All Media Is National,"
- Few people seem to care about state politics - They affect voters way more - Partisan Polarization and Impacts on Voting - The Transformation of Media Markets and how that affected voters' knowledge and participation levels - Americans are far more engaged with national than local or state - The media is the key engine of this nationalization - Odds of your vote meaning more in state and local elections is way higher! - Americans are moving away from outlets that would have state and local coverage to outlets that don't - Local TV news and newspapers bye
Religiosity is measured as
- Frequency of Religious Attendance - Frequency of Prayer - Asked "Is religion an important part of your daily life?" - It varies hugely and is great in comparing different religions and how they vote.
Other ways you learn political information
- Friends - Documentaries/Books/TV shows/movies
Abramowitz's "All Politics Is National: Using the Generic Ballot to Forecast the 2010 State Legislative Elections
- From the 2010 Midterm election - State legislative offices are very important in their respective states - Dramatic approaches between Democrats and Republicans - Redistricting also very important - Two main factors that caused GOP to gain a ton of seats: 1. Obama was a Democrat and very left leaning 2. Democrats won a ton of seats in 06 and 08 and now they are way overextended - This same pattern extends to State legislative elections - President's party lost 15 of 16 times, also 14 of 16 state legislatures - The Regression line shows the relationship between the Dems and GOP - NATIONAL POLITICAL FORCES INFLUENCE THE STATE ELECTIONS "BIGLY" - 2% is all that is needed to win more state chambers - All Politics is local is not true and Democrats got their clocks cleaned
Trende's "Does the GOP Have to Pass Immigration Reform?"
- GOP has to pass immigration reform to "survive" - GOP has actually done pretty well in elections with Hispanics and they can continue to grow this margin - What happens in 2016 when Obama isn't on the ballot? - blacks are voting at their natural rate at 13% - In 2016 it plummeted big league - Blacks vote for Democrats not always other blacks that are GOP - African Americans are against immigration reform and for protecting American workers - Populist economic message can win over blacks - 1.5 points every for years white voters are headed towards the GOP, which is a massive shift and consistent - There's actually been a slight shift to Republicans overall 25 states to 19 gain
Kennedy and Deane's "A basic question when reading a poll: Does it include or exclude nonvoters?
- GP (General Population of all adults) - RV (Registered adults) - LV (Likely voters) - In non election years, pollsters survey all adults - Voters as a group skew whiter and older than the general public - Surveying likely voters skews the votes - There's a clear support difference between RV-GP - Registered Voter polls exclude nearly 40% of the population - LV polls exclude even more. 55% voting rate among adults in prez election, and 33% in midterm election - For campaign pollsters, RV or LV provide the most accurate information about the mood and opinions of that subset of Americans who will actually cast a ballot in the next election - General population polls are better for the opinions of the entire country - Polls that exclude nonvoters paint a portrait of a public that is distinctly more active and ideological than all adults.
Noel's "Ideology is a Coalition"
- Ideologies will often clash over priorities - BLM wants to make racial equality the biggest Democrat issue - Bernie White voters have different issue priorities like health insurance and student debt. - "White liberals are white" - Conservatives have the same diversity tent of views - Ideaologues shape parties
Factors that explain/predict state legislative elections:
- Incumbent Governor Approval Endorsements - (Even local organization endorsements matter, due to the low information environment and low turnout) - If you want to win a state legislative election, be a person of your community. Candidate Mobilization/Ground Game: - State legislative elections have low turnout, especially when not paired with many higher office elections - Knocking on doors is a big deal and can outweight money Money - Matters more in state legislative elections because candidates can't always get access to money, depending on state and state House/Senate Name recognition - Few people outside of regular voters and grassroots activists know their state legislators, so this matters even more than in US House Elections Party ID: - A big factor when name recognition is low - The president's party loses seats in state legislative elections (Abramowitz) - Just like the US House and Senate - If it's a midterm election - # of seats being defended matters too - and the generic Congressional ballot!
Beams "Into the Blue"
- Long term demographics favor Democrats - Faster growing states are gaining House seats: like Texas, Georgia, and Arizona - Minority working class voters are increasing - Abramowitz says 1/3 of electorate in 2020 will be non white - Blue regions of red states is where the most growth is occurring which is why they may flip - GOP can gerrymander and keep stuff benefitting to win them, which is beneficial. - Political winds don't matter when "ground is shifting"
Gender Gaps
- Lots of persistent differences in women and men - In both social and political life
Notable campaign and media findings
- More candidate Google searches before a primary, increases candidate vote count - Both negative and positive ads can be effective, but maybe in different contexts - Increased facial recognition, increased votes (even from elevator pictures)
Darr's "The 2020 Candidates Aren't Matching Obama's Iowa Ground Game (Yet),
- Obama had 37 Iowa field Offices. All over. - Warren and Biden had 24 each, way less than Obama - This will probably hurt them against Trump - You need to cast a wider net to actually win delegates (so rural offices are good) - Candidates should open up offices in the 10 Obama only countries where today's campaigns have so far steered clear - Why they're avoiding it is a question that is still unanswered - lazy? Strategy changes? Or Efficiency?
Abramowitz and Saunders "Exploring the Bases of Partisanship in the American Electorate: Social Identity v Ideology"
- Partisan Hearts and Minds - Party ID is based on Identifying with a social group rather than a rational evaluation of the parties orientations - Social Identity Theory - Republicanism became desired by Southerners when their friends started switching parties - Most social characteristics are only weakly related to party ID, so the argument is not good. - Much larger rational component to party ID - Whites outside of the South becoming more exclusively GOP - Catholic advantage for Dems is gone - These are huge gains because there are so many Catholics - increasing clarity of ideological differences during Reagan and after has made it e easier for citizens to choose a party ID based on their true ideological orientations. - Conservatives aligned almost wholly with the Republican party. - Social identity theory: Party ID is based largely on membership in social groups - Most Jews are Democrats because of their liberal preferences, not social identity -
The Iron Triangle Model of Vote Choice
- Party ID - Ideology - Candidate Evaluations
Valence Issues /Valence
- Personal Characteristics (non-policy factors) unrelated to ideology or party - Ex: Incumbency, personal wealth, name recognition, charisma, intelligence, cool factor
Cohn's "After a Tough 2016, Many Pollsters Haven't Changed Anything"
- Polls broadly overstates Hillary's strength in Rust Belt strength - Broad near consensus among pollsters on the causes of error - Public pollsters haven't changed anything about their approach - Private pollsters have begun making the changes - Especially Democrats who were stunned - Private pollsters are doing something about education now - Weighting by education makes a big boost for Republicans - It ensures that less educated voters represent the appropriate share of the polls final estimate - Many individual public pollsters were happy with their results even though the industry as a whole got it so wrong - Most pollsters use voter file data rather than RDD - Even weighting by education wouldn't have been enough to give Trump the advantage - Civis says there's huge nonresponse challenges
Factors that matter for gubernatorial elections:
- Presidential Approval - State Economic Conditions - Incumbent Governor Approval
Gelman's "Why are Primaries Hard to Predict?"
- Prez elections are predictable - Change in per capita income - largely predictable from political ideology, degree of partisanship, and, again, recent economic conditions. - Not every election can be predicted ahead of time - State elections are often highly unbalanced in terms of resources - The candidates in a primary election are of the same political party and typically differ in only minor ways in their political positions, so it is easier for voters to change their opinions. Primary election campaigns can be highly unequal too - General election is mostly decided by the state of the economy
Issenberg's "The Mystery of the Puerto Rican Voter"
- Puerto Rico has massive turnout but when they get to the mainland US it plummets - Caravanas - MItt Romney tried one - Large groups of neighborhood mobs that encourage high turnout in Puerto Rico - PR turnout on average 10 points higher than the mainland - Cubans Conservative, PR's are swing voters - Of course, come up with an excuse for them registering as Republicans. - PR Governor endorsed Romney - PR elections = life is saturated with politics unlike in the US mainland - State Electoral Commission flatters resident - ELECTIONEERING WALKS CALLED CAMINATAS - Festivals are effetitve (Very skeptical example given) - Parties mean more votes
Campaigns use media to:
- Spread their message - Define their candidates - Define opponents - Raise issue awareness - Recruit volunteers and smaller donors
Turnout in Presidential Elections
- Stable during WW2 era - With the exception of 2012, it has risen since 1996 - 2008: 62.2% - 2016: 60.2% *Voter Eligible Population more important
Fundamentals of State and Local elections (BOLD)
- State elections have become more nationalized (BOLDED) - All politics is "not local" as Tip Oneill said in the 80s - % of Americans in Competitive Districts is low, and has been decreasing since the 70s.
Echelon Insights "One Republican Party, Different Democrat Parties"
- The various groups within the Democrat party seem to be more divided on flashpoint issues than GOP groups - Blacks are divided on some of the social issues and make up 1/5 of the Dem voting bloc although this is changing - White college educated Democrats are the group pulling the party and public discourse to the left on society and culture. - They are far further to the left than either Republicans who support Trump and Fox viewers - Primary voters are more ideological which we already know - Democrat donors are further left than GOP donors are to the right, very radical! - Shows how very liberal colleges and profs are producing very liberal voters.
Enten's Trump Probably Did Better With Latino Voters Than Romney Did
- Trump did better with Latinos than Romney did four years earlier - Can exit polls be trusted? - Latino Decisions, Segura and Barreto, exit polls we wrong - Trump "probably" did better and these clowns are wrong. - Republicans may be able to make Latino gains in the future if Trump did so well - Clinton lost 5 points with the Latino vote - Clinton underperformed Obama in counties where over 90 percent of the population is Hispanic - In heavily populated areas, she also did work. - In 04, Bush won 44 of the Latino vote.
Rose Garden Strategy
- Using the symbols/Trappings of the office to look Presidential and authoritative - A big advantage for incumbent Presidents
Candidate Evaluations
- Usually least important factor - These matter more for low information independent voters - EG (Trump candidate evaluation)
Cohort Effects
- You are more likely to have the Party ID of the party popular -> When you first become politically aware/active - Usually late teens/early college - These endure for life
Bottom up media (Social Media)
- weakens elite control over candidates such as Bloomberg, Sanders, Trump, and Cruz
Campaigns win by targeting
1, 2, or all of these 3 factors
4 Reasons why campaign effects are preditcable?
1. Campaigns are limited because many decide who to vote for before campaign (partisans especially) 2. Incumbents make campaign effects on voting often favor themselves due to incumbent advantages (what are they?) 3. Reactions to the election year economy become part of voter decisions during campaigns and influence views of the incumbent party candidate 4. Every campaign is highly competitive, and this competition narrows the lead of the candidate ahead at the start, though this is not the same as underdog momentum.
Losing the popular vote, but winning the election
1824: Jackson wins plurality of popular and electoral vote But loses the house vote to John Quincy Adams 1876: Tilden wins a majority of the popular vote to Hayes' 47.9% But loses a disputed electoral vote to Hayes by 1 electoral vote 1888: Cleveland wins he popular vote by .8% (though he won less than 50%) But loses the electoral vote to Benjamin Harrison 2000: Bush wins FL (by 537 votes) and the electoral vote Gore wins the popular vote (by .5%) but loses the election In practice, some of those who have won the election but the lost the Electoral College: Have still been close to winning the popular vote Elections like 1876 pose challenges to forecasting models and our understanding of elections
Proof of incumbency advantage
1992-2006: 90% of house incumbents are reelected every 2 years 65% of house incumbents receive 60% of 2 party vote 1946-2006 16% of House incumbents run unopposed 67% of House incumbents receive 60% of 2 party vote or run unopposed
Campbell's "The Seats in Trouble Forecast of the 2010 Elections to the U.S House"
2010 very good GOP year as predicted Large # of factors affect voting Democrats had a huge amount of seats to hold in red districts Partisan parity, polarization (Obama very liberal) which helped GOP Seats in trouble model highlighted that the Dems were overexposed Learning/toss-up seats are the most important Party registers gains where it's the easiest to do so President's party has gained seats in only 1 of 32 elections between 1870 and 1994 (this is changing with polarization) 65% approval is the neutral line of a President to gain/hold seats GOP Going to wreck the Dems
Ideology/Issue Positions
2nd Most Important Factor (very related to party ID today)
Caravanas
A mob of cars in Puerto Rico that drives around supporting a particular candidate;.;
Forecast Timing
A model predicting tomorrow's election is mostly useless • Unless it's a close election • Or head to head polling varies a lot (2012) o Nate Silver's 538 model correctly forecast an Obama win in 2012 o Useful informal the day before an election o Because of varying poll results, especially with state polls and national polls conflicting o Even accounting for wide polling variation in 2012 o Romney did better national polls o State pols were more accurate A model predicting an election 22 months in advance is mostly useless • 22 months is a lifetime in politics • lots of political and economic change can occur in 22 months • multiple congressional sessions, candidates haven't declared yet, no endorsements, etc.
Ideal Time to Forecast
About 15 months to 5 months before an election Candidates and then nominees are known Which means you can account for the effects of: • Ideology of candidates • Short/long primaries • Election unclear economy • Why? We can see fundraising, campaign dynamics and an idea of how divided the field is. • General state of the economy
Turnout is much higher when
An election is close
Targeted Ads
Are cheaper and can be effective (You can avoid messaging the wrong voters!)
Online Polls
Are effectively opt-in So they're not random, except within the interviewer's pool of respondents They don't measure the same group of people as telephone polls
Marriage Gap is not
As strong as the gun household gap in voting.
Omitted Variable Bias
Assuming that a relationship you've observed takes into account all relevant factors, but taking into account other factors changes the relationship. - Education and policy views, education and PID
Retirement Slump
Average drop in party vote share from previous election when incumbent leaves the seat and seat is open
White voters
Becoming more GOP, and conservative
Nilsen's "Can Joe Biden win the nomination even if he loses Iowa and New Hampshire?"
Biden thinks can win w/o Iowa and NH Must win Obama's nonwhite voters No single state is a must-win No candidate in the modern primary cycle has won the nomination without coming in first or second in New Hampshire or Iowa The election headlines/who overperformed are the most important things Butt won Iowa Biden will collapse if he does poor in NH Superdelegates power lessened 15% threshold Bernie focused heavily on CA (****ing huge)
Income Relationship
Bivariate Relationship (HIgher Income = More Republican) although this is decreasing
Gender Gap in Partisanship
Bivariate Relationship - women are more liberal and more Dem
The Seats-in-Trouble Forecasts of the 2018 Midterm Congressional Election
Blue Wave Year Dems idiots Has been wrong some years like 2016 Toss up middle states usually favors GOP Democrats should gain 44 seats based on the model and take control of the House Correctly predicted that GOP would hold the senate
Media Coverage
Book argues that the party establishment can drive media coverage Which can determine winners Because name recognition and perceptions are important for candidate support Especially for less visible but otherwise strong candidates Though media coverage is the least important of the 3 parts of the invisible primary The book (& receptive analysts) focus those 3 parts of the invisible primary But there are other factors, theoretically part of it, which matter
96.5 of Americans in 2014: (BOLD)
Both of their incumbent state legislators win re-election.
Telephone polls don't reach everyone
But for now, can be randomized for a selected population... (KIND OF) Even Rasumussen uses an online panel now
Theory of Minimal Effects
Campaign Effects are minimal because: - Both the positive effects of new info and - negative effects of prior info on vote decisions Come from a common source: Political Interest Increase in one is correlated with increase in the other. Which results in little or no change in voting from campaigns.
Factors that explain/predict gubernatorial and state legislative lections:
Charlie Cook's PVI (Partisan Voting Index)
At what point does the establishment majority need to decide?
Charts on pgs. 176-177 show lots of variation in # of governors endorsing by IA caucus If the establishment "decides" after the nominee is already apparent It doesn't really count as party influence Once candidates start to win primaries, the establishment may no longer be able to stop them
- C + D in the equation
Costs of voting + participatory benefits of voting (like sense of Duty)
Women and Men Have
Distinct Educational Experiences, different degrees and different workspace gender makeups
Ferguson's Advice
Do both for your campaign Because differences in results are hard to anticipate
Noel's "What can 2016 tell us about 2020?"
Donald Trump won the majority share of his party Narrow but Intense following can come in first with a rather small plurality of the vote There is no Donald Trump running for the Democratic nomination Current Tea Party of the Democrats like Ocasio and Tlaib Biden as Bush, Bernie as Cruz Trump showed how limited the Party Decides power can be
The traditional journalist view of campaigns
EVERYTHING MATTERS, but this is fake news.
Party ID and Issue Positions Cause Changes in
Each Other
Primary
Each party allocates delegates to each state Delegates cast ballots at the Republican and Democratic National Conventions for the party's nominee Each party awards delegate differently The GOP gives bonuses to states based on how many state and national GOP officeholders it has The Democrats give bonuses to states if they schedule their primaries later But more populated states get more for both As do states that have repeatedly voted for that party's presidential candidates Presidential primaries are elections run by that state's government
Influential factors
Economy, public opinion and foreign policy are all influential
Religion and Religiosity
Especially for Hispanic Americans
Farley's Law
FDR's campaign manager said most elections are decided before the campaign begins. Most voters decide who to vote for early in election years and partisanship decides most of this.
Silver's (FiveThirtyEight Senate Forecast: GOP is Slight Fav - ONLY READ UNDER OVERVIEW OF OUR METHODOLOGY)
GOP now slight favorites to win at least six seats and capture the chamber The national environment: the single best measurement of the National political environment is the generic congressional ballot. Also, look at Obama's approval ratings Candidate quality (fundraising totals, ideology ratings) State partisanship (Partisan Voting Index, PVI) Incumbency is huge - they win the overwhelming majority of races Head to head polls Insults Steve Rasmussen
Rakich's "Our Best Tool for Predicting Midterm Elections Works in Presidential Years too"
Generic congressional ballot predicted a wave The political environment still favors democrats (not anymore) Generic Congressional Ballot is one of the most accurate predictors of who will get the most votes for Congress in a midterm election. It's just as Predictive during presidential elections Generic Ballot Polling margin and national House popular vote margin Partisanship is the dominant force in today's politics, these days, people straight-ticket vote.
Urban Areas
Getting More Democrat
Rural Areas
Getting More Republican
Later media coverage
Greatly amplifies the effects of Political ads, rallies, protests, speeches, and debates (Secondary media coverage)
The South
Has much higher religiosity
Campbells Argument
Has weaknesses that political professionals would point out.
State elections
Have become nationalized
Americans are divided by religiosity
Higher religiosity is more GOP, GOP Voting, and more conservative
PVI (Partisan Voting Index) (BOLD)
How much a House district's/state's Presidential voting patterns differ from the national average. State/District party ID. Basically how far left or right is that district compared to the average.
Vision of good society
How you want society to look
Rational Choice Thoery
Ideology and Issue Views When you realize there's a liberal Democratic Party and a Conservative Republican Party ---> You pick which one's closest to your views
Broadly targeted TV and radio advertising is very
Imprecise. It misses so MANY People today. Except for a few events (Super Bowl, SOTU) - Americans rarely are all watching/listening to the same thing
1 very politically active person in a group
Increases the activity of the whole group
Cambells View
Is a balance between the political science AND journalist view.
Wedge Issues
Issues which some % of the other party's supporters agree with your party on
Religiosity is more important for politics than religion
It tells how religious you are, your devoutnes, regardless of religion
Casual Relationships (X->Y)
Living in a rural/urban area makes you more Republican/Democratic - Why? - City Dwellers are more liberal in most other countries as well.
Establishment majority?
Makes sense, but isn't specifically defined either How much of a majority is necessary? 51%? 67%? 75% 90% Insuffiently large establishment majorities supporting a candidate by July or year prior
Incumbency Effects for State Legislative Elections
Matter LESS than they used to
How much is incumbency worth in House elections?
Maybe 5-8 percent of the vote since 1990 Which can be a lot, though not that many districts are competitive Less in Senate Elections
Seniors are
More GOP as of recent
Economist's "Prediction 2016: How Jesse Jackson Inadvertently Revived Political Betting"
Most polls had Romney leading by a point Betting markets had considered the incumbent a heavy favorite Started in 1820 and continued to grow in 1916 1926 George Gallup killed the electoral markets Jesse Jackson's Michigan Primary Victory shocked many in 1988 Caused University of Iowa professors to set up the Iowa Political Stock Market Iowa Electronic Markets 2004 Intrade came around PredictIt is the New Zealand new alternative - bets capped at 850 wrong about Trump
Key factors in congressional elections
Name recognition Whether voters know who a candidate on the ballot is When voters know incumbent but not challenger, they may vote for the party's candidate The reverse is also true, but is less likely to occur Voters know winning challengers Other factors in Congressional elections: Voter turnout in Congressional (off year elections: Usually around 40 percent of eligible voters
Are Republicans and Dems divided by religion?
No, they are not. Voting differences between Protestants and Catholics are small, they are 70% of all voters.
How many voters decide within 2 weeks of Election Day
ONLY 10-25% even though this is when the most people are paying attention.
Which Groups of Americans Don't Vote?
P: People in uncompetitive states B: Anyone who has difficulty telling the difference between the candidates (The poorly educated etc) C: The very poor, people who live in places where it is difficult to vote, naturalized citizens. D: Those without political friends, people in communities that don't care about politics.
Douglas Hibbs' "Bread and Peace Voting in US Presidential Elections: What Impact Rising Inequality?
People are promised more than they can be promised The question of bread and the question of Peace Romney, not Obama's opponent - the economy was Killing Vietnamese soldiers at a rate of 10 to 1 but people were outraged at the killings According to the Bread and Peace model elections are a sequence of referendums on the incumbent party's record. (1) Weighted average growth of per capita income over the term (2) Cumulative US military fatalities owing to unprovoked deployments of US troops in foreign wars Aggregate votes and voting can be often explained by these measures alone Factor 1: Bread - Economic preference is the dominant factor (voting in retrospective) Growth of per capita real disposable income Factor 2: Peace - Presidents inheriting a foreign war get a 1 term grace period Didn't affect Bush in 04 Fatalities don't affect Congressional outcomes The Bread and Peace model is designed to explain presidential elections. NOT predict them. Polls are not used Attitudinal variables are endogenous Johnson in 1964 due to Bread, 1964 and 1980 big Breadwinners 1956 and 1972 are Peace
Selection Effect
People who choose to live in urban/rural areas as adults are already more Democrat/Republican Living in a rural/urban area does NOT make you more Republican or Democrat
As media coverage goes up
Poll support goes up (Ben Carson Fall 2015)
You will vote if V is
Positive, and not when it is negative.
(P*B) in the equation
Probability of being a pivotal voter * difference in Benefits you recieve from the 2 candidates
3 Theories of how people change parties
Rational Choice, Social Identity, Negative Partisanship
Masket's "The 2020 Invisible Primary in light of 2016"
Right now is the invisible primary 2016 Theory was totally spot-on for Dems with Hillary Democrats are still preferring traditionally credentialed candidates (career pols) 5 important potential outcomes for the eventual nomination
Theory of Post-Materialism
Ronald Inglehart - People with post-material values don't trust MNCs or government Value local community
The Theory of Post Materialism
Shows why younger people in the West are more liberal because they've never had economic or physical security threats like past generations.
Men are
Slightly more informed, slightly more interested in politics
Asian Voters
Somewhat Dem
Coattail Effect
Successful candidates at the top of the ballot pull up some of their party's candidates into office with them. Those who vote for one party's candidate in a Presidential election: Increases (^) change of voting for that party in Congress
Coattail effects also matter
Successful candidates at top of ticket (usually Presidents) "Pull some of their party's candidates into office along with them" Those who vote for one party's candidate in a Presidential lection: Are more likely to vote for that party's Congress candidate Even accounting for individual partisanship and candidate evaluations
Hispanic Men
Tend to be more GOP, especially in TX and FL But more Dem in CA and NY
Party ID
The Most important factor
Relationship Between Congressional and State Legislative elections
The correlation is almost an even 1.
Theory of Minimal Effects Continued:
The most attentive to campaign info are actually the least likely to change their votes. The least attentive to campaign info are the most receptive to changing their votes.
Norpoth's "The Electoral Cycle"
The most important thing to factor is the two-term rule A trigonometric function cannot model human behavior Can we actually predict cyclical election cycles? When the White House party has been in for 1 term, the odds are 6/7 that they will win re-election regardless of party After two or more terms, they now can lose 6/7 times The two outliers were more or less successors (Ford, Bush) The standard error is tiny on this cycle model
Campbell attempts to answer, do presidential campaigns matter?
The traditional political science view is NO
Silver's "15 Percent Is Not a Magic Number for Primary Delegates"
They have both state and district level delegates Proportionally distributed above 15% 35% are state-level and 65% are district level Still miss out on districts at 15% and get some too The safest bet for a candidate is to win 20% in a district
Social Benefits and Social Pressures
Things you may experience during an election. Social benefits are the benefits you get socially from voting, like your friend thinking you're cool. Social pressure is your friends or family pressuring you to vote, for example.
Abramowitz's "Forecasting the 2016 Presidential Election: Will Time for Change Mean Time for Trump?"
Time for Change Forecasting Model Incumbent President's Approval Rating in the Gallup Poll at Midyear, The Growth Rate of the Real GDP in Q2 of an election year, and whether or not the President is an incumbent Very, very difficult to win the White House after having been in for 2 terms 55 v 49 in second v third term (party) Time for Change predicted Trump to win correctly
Silver's "We're Tracking 2020 Presidential Endorsements - Here's Why they Probably Still Matter"
Trump had no endorsements in the invisible primary until he started winning states Despite 2016, the Party Mostly decides Minimally acceptable candidates are usually always chosen 9 out of 11 times the candidate with the most endorsements has won the primary McGovern Fraser reform did little to stop people from agreeing with party elites National polls do comparably well to endorsements (Chart) 11 out of the 16 nomination races, pre-polling and endorsement winners were the same
The Rational Choice Theory of Voting (Rider and Ordeshook 1968)
V = (P*B) - C + D -
Full Equation of V=(P*B)-C+D
V = Value from Voting (P*B) = Probability of being a pivotal voter * Difference in benefits you receive from 2 candidates - C (Costs of Voting) + D (Participatory benefits, like Sense of Duty) You win if V is positive, and don't if it is negative.
Which Groups of Americans Vote?
V=(P*B)-C+D P: People in battleground states B: Anyone who sees benefits in one candidate over the other - Politics followers, the well educated, the ideological/issue-oriented/strong partisans, the devoutly religious, senior citizens C: High income, people who live in easy voting states D: Married people, those with politically active friends, civic education course takers
Party ID changes
VERY SLOWLY, so even small changes matter
V in the equation
Value from voting
Black Voters
Very Democrat, but decreasing very slowly
Candidate Evaluations (Iron Triangle)
Voter perceptions of politicians - 3rd factor (Usually most important) in iron triangle model - Ex: Job Approval, Most Trusted on Economy, Best Experience, Can Bring Change To Washington, ETC - Often most interesting - Include Valence Issues
Negative Partisanship
Voters form strong loyalties based on loathing for the opposing party This is instead of: - It being inherited - Based on ideology - Being of the wrong party matters A LOT for voting - Affection for a party matters less than hatred for other party
Ideologies are Coalitions
Voters prioritize issues differently (What needs to be done first)
Masket's "Who gets the nomination will tell us a lot about how strong the party is"
Who wins strength meter Booker Harris Biden - strong , Klobo and Castro somewhat Warren Butt Bernie Bloomberg Gabbard - Mixed all the way to party is dead The date was written (November 12th, 2019)
Gender Gap in turnout
Women vote slightly more than men, 4% more women vote Why? Men are in prison and felons.
Social Identity Theory
You are often socialized into a party by: - Family and Friends Even if not, at some point you ask, which party has people like me, and pick which one fits demographically. - Your grandfather was a proud Republican!
What are the Benefits you can recieve if your candidate wins?
You have to decide as a voter.
Ideology
a cohesive set of ideas and beliefs that form a general philosophy about the role of government in society Gives you those views on particular issues. - Reflects how someone sees the political and social world - Related to someone's vision of the good society
The Party Decides
basic argument: Despite party reforms introducing voter selection of convention delegates in 1972, Intended to give voters more control over Presidential nominees The party establishment gets the nominees they want
State elections have
become more nationalized
Multivariate
involving more than two variables
The 538 model
provided evidence that state polls are better predictors of presidential election results than national pols A day before forecast can be useful, but not usually 538: Lots of issues may affect the election this is a more typical journalist view but points out )correctly) that more issues besides the economy affect elections
bivariate relationship
relationship between two variables
Candidate Evaluations are
the most malleable of the 3 iron triangle factors Campaigns spend a hefty fortune on these.
rational ignorance
the state of being uninformed about politics because of the cost in time and energy - voters know enough but not too much.
Marriage Gap
the tendency for married people to hold political opinions that differ from those of people who have never married - More likely to vote - More likely to be Republican, and Vote GOP - Married People are more likely to be PID and vote Republican
Hard to know
whether online polls or TV polls are better
Super Delegates
• Aka unpledged or automatic delegates • Democratic Party: • Current or former party leaders or elected officials • All democratic governors, leaders, and part people • 3 Republican National Committee members from each state • 7% or republican delegates • And 16% for democratic delegates • DNC ones aren't bound to vote for any candidate • GOP: some are bound to state's election results
Field organization/Ground game
• Building offices in states, organizing volunteers, and leveraging local networks to persuade voters • Matter a lot in primaries, and especially in caucuses
Winner take all nature
• Disadvantages 3rd parties because people don't like "wasting" their vote • & because the closely divided Democrats & GOP have strong incentives and ability to co-opt them into their coalitions
Hibb's (Bread and Peace)
• Economy • Foreign Policy • Are all that matters for Presidential election outcomes • Though he does hedge with "due to unprovoked, hostile deployments" part of the Peace component
Casework
• Helping constituents deal with the federal bureaucracy • Ex. VA (Veterans Affairs), HUD (Housing and Urban Development), USCIS (united States Citizenship & Immigration Services) • Reaches across party lines • MCs spend a lot of resources on it, & it helps them get re-elected • Constituents LOVE it (my example) • Since government bigger than it used ot be, more casework to do
Primary voters are generally:
• Informed • Ideological and issue oriented • Liberal and Conservative • Politically active • Form the core of campaign's volunteers (free labor) and small to medium donations
WHYA RE BETTING MARKETS BACK?
• Internet and easier commerce • More legality (internet allows ease of operation where legal too) • Declining accuracy of polling increases, while profit from expertise or insider info increases again
How far in advance?
• Maybe As close to the election as possible, • Some people may stop paying attention though • Reasons to do forecasting far in advance and reason not to • You want to hit a sweet spot
Money and Incumbency Advantage
• Money is important, but mostly for challengers, who need it to get their message out • On average, challengers need to spend $600,000 to have a 20% chance of winning • Incumbents can get money when they need it • Incumbents pend heavily when they are more likely to lose • SO when they do, predict they might lose! • Increase in spending by challenging=increase in chance of winning • Increase in spending by incumbent= decrease in chance of winning • Challenger Dave Brat: $200,000 • Eric Cantor: $5 million
The invisible primary
• Often begins 2 years before the IA caucuses • Candidates commission polls, start raising money, hire staff, seek endorsements, & build their profile in the media • Called invisible because most of it isn't directly observable • Not that well reported in the press outside of some polling and fundraising • And most primary voters aren't yet paying attention
Why did Presidential election markets decline between 1940 and 2000?
• Polling becomes an effective science • Decrease: News interest in markets as a gauge of the race • Polling accuracy publicly reduces election uncertainty in markets • Uncertainty: you make far less money from expertise or insider info • Ironically, politics also caused the decline in election markets • Election markets became illegal in NY (industry center) • And couldn't compete against govt. sponsored gambling alternates • Economist graph doesn't show election markets disappeared 1940-2000 • Underground, they just weren't in the news • 1964, Don Sal and a daring professor
Betting Markets
• Prediction 2016 article • Betting markets have a long, successful history in America • In the pre-polling era, 1884-1940 • Markets call 11 of 12 presidential elections right • Exception: 1916 (Wilson defeats Hughes in toss up) • 2000: online betting became huge
Multi-Ballots
• Presidential candidates nominated after multiple ballots less likely to win
The electoral college ensures that:
• Presidential candidates treat states as individual political units • And campaign on state by state basis • Especially in purple or swing states • Rather than just in big cities and media cent • Which is what would happen in a national popular vote
9 big factors
• Presidential job approval • Generic Congressional ballot/polling for particular elections • # of quality challengers running • Economy/war • # of seats held by President's party • # of swing or exposed districts held by president's party • Public Opinion for liberal and conservative government • State PVI (average of outperformance0
Fundamentals of primary elections
• Presidential primaries last from February 1 to mid-June in 2016 • The delegate leader when 50% of the delegates are allocated (March 8): • Is often able to clinch the nomination by the time 75% of delegates are allocated (April 26) • Linke in 2008 and 2012 • In practice: the nominee is decided by late April • No election since 1976 has had a nominee undecided before the party's National Convention
Electoral College
• Prevents direct election of the president of the public and the election of a demagogue • Each candidate chooses electors who cast their electoral votes for a candidate • # of electors for each states=# of house and senate members it has • Least populated states have 3, DC also has 3 • You must get a majority of electoral votes to win, or the election is decided by the House • Electors: each chosen by popular vote in their states • Each candidate has electors attached to him or her • Electors of the candidate who wins statewide are chosen • Except in NE Maine
Quality challengers are strategic
• Quality candidates • For political scientists, those who have held prior elective office • Incumbents are by definition quality candidates • Unless candidates are vulnerable, they draw weak challengers
Electoral College reform ideas
• Reduce/increase # of electoral votes for less populated states • Allocate electors nationwide by Congressional District • Increase winning condition from electoral vote majority to super majority • Change most states from winner take-all to proportional allocation • All of these provide advantages/disadvantages to each party
WHY is the incumbency advantage bigger for reps than senators?
• Reps more connected and in touch with smaller constituencies • Senators are more prominent in media for controversial votes • Senate challengers more well known and get more media attention
Incumbency Advantage
• Sitting MCs (members of congress) have an advantage over challengers in elections • Bigger in the past (1970s/1980s especially) than now • But still matters a lot • Ex. 2006 Congressional Elections • Democrats gain 31 House seats • But 90% of GOP incumbents are re-elected
Predicting congressional elections
• States and House Districts vary a lot in character: • Geography, population, economy, income, ethnicity, age, party support • Which makes predicting/explaining Congressional elections harder than Presidential elections • Especially individual elections • & when 1 to 1 comparing states/districts • But we still know a lot
Frontloading
• States want to move their primary earlier to get attention from the media • Both the DNC and the RNC now heavily punish states for doing this • By considerably reducing their number of delegates • Most (75%) of Presidential primaries are primary elections, not caucuses
Technology and Data
• The ability to target micro targeted persuasive and turnout matters • As does the ability to track support, identify supporters, and best use resources • And there is lots of variation in candidate data abilities recently
The endorsement primary
• The competition to claim endorsements from the establishment • Most important public signal of establishment support • Candidates who win the majority of establishment endorsements often win
Three parts to invisible primary
• The endorsement primary • The Money primary • Media Coverage
The money primary
• The race to win campaign donations & today super PAC support • Super PAC $ matter else than campaign donations • Donations matter because: • They signal future strong support • And campaign organization strength • Though the endorsement primary is 2.5 time as good a predictor as the money primary!
COngressional elections ar emore nationalized
• They're more determined by national factors like: how many seats the president's party holds (more seats, ore exposed) • Change in income per capita in year before election (and other economic variables) • Presidential Job approval
WHY ARE election markets a good predictor of elections?
• Traders have to back up their opinions with cash • Which commentators & journalists rarely have to do • If you're going to bet, you at least think you know more than the average person • So the people who bet are the most informed/intelligent • Markets can update constantly & more quickly than polling o React immediately after and even during debates • Once you understand how they work, in some ways easier to understand than polls o No stats, no hidden assumptions • Can incorporate insider info o Upcoming endorsement, ground game info, private polling, scandal knowledge
Why has Hibbs not done so well recently?
• While he's got the economy and foreign policy fundamentals of elections: • He's left out incumbent job approval or public attitudes of any kind. • Incumbent President Party Share of 2 party vote, 1948-2016, predicted by: • Incumbent President Job approval, 1st January Poll of Election Year • Inflation Rate, election year • Consumer Confidence, August of Election year • Yes/No whether out party candidate is more ideologically extreme than in party candidate • Election Year GDP growth (2% for this year) • Change in defense spending as % of election year budget since election year
