Latest / Politics with Paul / Politics with Paul Podcast #8 - How Do Election Forecasters Make Their Predictions?
Transcript
- 0:01It's time for politics.
- 0:02Welcome back to the politics with Paul podcast with your host, Paul Reeder.
- 0:07Today, we're going to talk about election predictions.
- 0:11So we're in the home stretch of the presidential election season for 2024, and in
- 0:17such a homestretch or the homestretch for any major election, you're going to see increasing
- 0:23attention paid to these predictions of their results, Which presidential candidate is going
- 0:29to win Nevada or North Carolina?
- 0:33How about who's going to win the governor's race in Washington, for example?
- 0:39So many statewide elections or US house district elections are seen as clearly favoring
- 0:45either the Democratic party or the Republican party.
- 0:48Others are seen as moderately competitive, but probably leaning towards somebody else.
- 0:53Some are seen as probably leaning towards this other candidate, but it's anybody's game.
- 0:57And then others are considered complete toss ups where
- 1:00who knows who's going to win those those elections.
- 1:04But how can forecasters be so sure about a lot of these contests and other ones not be so sure
- 1:12about them.
- 1:14How do they come up with these conclusions?
- 1:16Well, today, that's what we're gonna talk about.
- 1:19So it makes sense to first start off with big name forecasters and their formulas.
- 1:25Organizational examples include the Cook Political Report, Larry Sabato's crystal ball.
- 1:30He is a professor at the University of Virginia and runs the Center For Politics, or maybe he
- 1:36doesn't run it anymore, but I believe he still does.
- 1:39Inside elections with Nathan Gonzales, see analysis, split ticket, elections daily, and
- 1:47these forecasters have both free and paid content explaining the reasoning for some of
- 1:51these races.
- 1:52They don't necessarily say that this party
- 1:55is going to win x state or x district in the US house.
- 2:00They just say how likely it is that that will happen, although they may say in their more
- 2:05detailed explanation for why they make these ratings.
- 2:08They might say, this candidate basically has it in the bag.
- 2:12Then there are individual examples like Nate Silver and Alan Lichtman.
- 2:17Silver was a statistician who founded the website 538, which blogged about statistics and
- 2:22polling, sports, and other topics.
- 2:24Since then, Silver has left, and 538 has ceased operations.
- 2:29Lichtman is a history professor at the at American University over in Washington, DC, and
- 2:36he has correctly predicted 9 of the last 10 elections except for 2,000, which was a pretty
- 2:42wacky election.
- 2:45So these forecasters don't necessarily share their exact formulas.
- 2:48They're not gonna say, okay.
- 2:50We put this much weight on this factor.
- 2:54This factor over here, we put less weight on.
- 2:56They're not gonna necessarily share those kinds of things, but they might be more open about
- 3:01what their formulas are.
- 3:02So, for example, Lichtman even had a book, that details 13 keys to presidential elections, and
- 3:10they include both objective and subjective factors like economic predictors and a
- 3:15candidate's charisma.
- 3:17Candidate charisma is kind of more of the subjective factor, even though some people you
- 3:23could just tell they just have that charisma.
- 3:26Whether you like them or you don't like them, that's not the point.
- 3:29And for Lichtman's formula, if 8 or more keys are true for the incumbent party, that
- 3:36incumbent will win the election.
- 3:38Technically, Kamala Harris is not the incumbent.
- 3:41Joe Biden is, but it's the incumbent party.
- 3:44So in this case, Democratic party is the incumbent for the presidential election.
- 3:49If she has a plus keys favoring her, then she should win the election, and Lichtman has come
- 3:57out and said that he believes Tamela Harris is going to win this election.
- 4:01So we'll see how that works out.
- 4:03But there's general agreement, between the forecasters on most federal contests and even
- 4:10governor's contests too.
- 4:12Typically, you see this in the what are considered the safe seats for democrats or the
- 4:17safe seats for republicans.
- 4:21A little bit more variation in the more competitive elections that are going on.
- 4:29Some people might see something as a toss-up.
- 4:31Others see it slightly favoring the Republican party or slightly favoring the Democratic
- 4:36party, but you see far fewer predictions made for the state level.
- 4:42Governor is an exception, but you don't really see it for elections such as attorney general
- 4:49or auditors and treasurers, and you you don't see it very often for state house or state
- 4:55senate.
- 4:56An exception in that regard is CN analysis, and the reason why is, well, there's more coverage
- 5:01and attention paid to these federal elections and governor, not so much all these other state
- 5:06level elections, and certainly not for local races, and these factors for the state level
- 5:12offices are often much more obscure and or less available for certain state level contests,
- 5:19such as for state senate.
- 5:22But you do see them every once in a while, but the predictions are primarily going to pertain
- 5:29to US house, US senate president, and governor.
- 5:34Now as for these levels of competitiveness, you see some similar terminology and levels of
- 5:40distinctions.
- 5:41The exact criteria for how to determine how likely the democrat or republican is to win
- 5:47a contest, again, that's that's just depends on the the predictor.
- 5:54And then the actual naming convention for these is also a
- 5:59little bit different too.
- 6:00So, for example, inside elections has the toss-up, then there's a hilt, then there's a
- 6:06lean, then there's likely, and then there's solid.
- 6:09Others like, Larry Sabato, I believe, uses yeah.
- 6:12Larry Sabato uses the designation of safe instead of solid, but it basically means the
- 6:18same thing.
- 6:18So it's it's useful to think of levels of competitiveness in terms of numerical levels.
- 6:23So a level 0 would be a toss-up.
- 6:26Who knows who's going to win this election?
- 6:28Level 1 is slightly favoring the Democratic or Republican Party.
- 6:32Level 2 is there's a fair chance that the Democrat or Republican are going to win, but
- 6:37it's they don't have it in the bag.
- 6:39And level 3 is no.
- 6:40This is basically a foregone conclusion.
- 6:43Now CNN analysis and inside elections have more tiers than the other election forecasters do.
- 6:50But if you're gonna compare between Sutz and Sabado and inside elections, see analysis and
- 6:57split ticket and elections daily, you can kinda think of it in those terms, those numerical
- 7:02levels 0, 1, 2, 3, level 1 republican, level 1 democrat, that sort of thing.
- 7:08So what are the kinds of factors that we look for when we're making these predictions?
- 7:13So there's a whole host of factors, and some forecasters put more weight on some factors
- 7:18versus others.
- 7:20Some may include factors that the others don't.
- 7:24Again, we don't necessarily know the exact formulas that these predictors do or and some
- 7:30of them are more open about how exactly they do it, but they'll still have content where they
- 7:35talk about, this has this has become a more likely democratic race, or this has become a
- 7:41more likely republican race because of x y z reasons.
- 7:44And you're gonna see that a lot more as we approach the November election.
- 7:49And these factors include both objective and subjective aspects.
- 7:54And then, of course, you've also got short term and long term forces, short term being those
- 7:59economic factors like unemployment or or wages, long term being previous electoral history both
- 8:06for individual voters and for overall contest.
- 8:09So how has have these voters voted in the past, and how these overall elections played out in
- 8:17previous iterations.
- 8:18So let's go take a deep dive into some of these specific factors.
- 8:23Not necessarily a comprehensive list, but just an assortment of some of the really big ones
- 8:29that are going to be considered when you make these predictions.
- 8:331st factor, previous history of the same elections results.
- 8:37History is one of, if not the strongest indicator of who wins the next election.
- 8:42So we're taking a specific contest like a governor in Maryland or or Washington, we said
- 8:51before, Kansas, whatever the case may be.
- 8:54If one party is consistently winning that contest and they're winning it handily, they'll
- 8:59probably win it again.
- 9:00For example, Democrats have won 9 straight US senate races in Michigan, but the margins have
- 9:05decreased in each of the last four times.
- 9:082020 was very close.
- 9:10It was was within 2 percentage points, percentage points being if you look at the
- 9:15numbers, there were 2 points between them.
- 9:17So, for example, if the first candidate or the winning candidate won 46%, and then the 2nd
- 9:23place candidate won 44%, that's a 2 percentage point difference.
- 9:26But big swings in margins of victory are possible.
- 9:29You don't necessarily see them that often, but it could happen.
- 9:33But generally speaking, if one party is just winning over and over and over again and it's
- 9:39not even close, and over time, the margins of victory haven't been getting
- 9:43closer, probably gonna be won by that same party again.
- 9:48Now what
- 9:49if it's a race that can happen in both midterm and presidential election years?
- 9:53Why is that even important?
- 9:55So we're talking namely US senate and US house because US house is 2 year terms, US senate is
- 10:016 year terms.
- 10:03The reason why is you you wanna compare similar years for that contest to see the most similar
- 10:10turnouts.
- 10:11Presidential election turnouts are more similar to other presidential elections, and midterm
- 10:16election turnouts are more similar to midterm elections.
- 10:19It's not that you can't compare the 2, but you might wanna base your prediction a bit more on,
- 10:27in this case, let's talk about 2024 being a presidential election.
- 10:31If you're talking about a US senate race that is happening if a US senate election happened
- 10:37in I don't I'm I can't remember for certain if it's actually happening this year, but let's
- 10:43say Oklahoma, and it also happened in 2020.
- 10:49You may wanna compare those 2 as opposed to if another state's US senate election was is gonna
- 10:55be happening this year, and then it had another US senate election in 2022, just because a lot
- 11:02of these presidential election voters do not come out for a midterm election.
- 11:07They're they're strictly presidential election voting.
- 11:10So, again, it's not that you can't compare the 2, but if you're basing presidential election
- 11:15forecasts or presidential election year forecasts, so not necessarily presidential
- 11:20election, but ones happening in a presidential election year, and you're basing it on a
- 11:26midterm election's results.
- 11:28It's not that you can't do that, but you gotta take it with a big grain of salt.
- 11:32Next factor, primary results, turnout.
- 11:35So primaries already get much lower turnout than a general does.
- 11:40But if the primary turnout for one party is particularly lower than the other party in
- 11:46terms of turnout rates, so there's 30% Republicans that come out but only 15%
- 11:52Democrats, that may be a bad sign for that disadvantaged party.
- 11:56Or if it's higher primary turnout overall, probably gonna mean more turnout for that
- 12:02general election.
- 12:03And if, again, if that one party is higher in primary turnout than they usually are, that
- 12:08might indicate that that party is more energized.
- 12:12And you also got to account for, in an election, if there's more Republicans or if
- 12:16there's more Democrats.
- 12:18So even if one party has a particularly high rate of turnout, let's say Republicans have 15
- 12:26no.
- 12:26Let's say they have 25 percent, turnout and Democrats have 15% turnout, but there's, like,
- 12:33a 1000000 more Democrats in whatever district or or state than there are Republicans, that
- 12:41difference in turnout rate might not matter because of the sheer number of, sheer advantage
- 12:48that Democrats have in total number of voters.
- 12:53Same thing if if it was Republicans who had the advantage in total numbers.
- 13:00Now there is some debate as to whether higher turnout elections favor Democrats.
- 13:06The fact is most democratic voter demographics tend to be given toward lower voter turnout.
- 13:12Republicans are typically wider, they're typically higher income, those those are
- 13:18associated with higher rates
- 13:20of turnout.
- 13:22But what's debatable is that if all else is equal, if overall turnout is higher, the
- 13:28democrats benefit.
- 13:29That's the idea.
- 13:31Some people still operate on that assumption.
- 13:33So if they see that there's higher turnout overall in a primary, they might suspect that
- 13:40the Democrat has an advantage.
- 13:41Again, that's assuming that all else is equal, like if there's an equal number of Republicans
- 13:48and Democrats.
- 13:50Rarely plays out like that, rarely is so clean, but that idea that higher turnout rates per,
- 13:57that they favor democrats overall, that's still a debate.
- 14:02Next factor, polling.
- 14:04So there's lots of misunderstanding about what polls really are and how they work.
- 14:09A poll is a sample estimating overall support for a candidate or ballot measure in a snapshot
- 14:17in time.
- 14:17Because polls are a sample of a population, it is not a perfect representation of the overall
- 14:22population because you can't get everybody's opinion on a candidate or a ballot measure.
- 14:29And for some people, their opinion might change over time.
- 14:32But mathematically, what may seem like a small sample size to the general population can
- 14:37actually be mathematically representative.
- 14:40It's a whole formula that I won't bog you down with, but as much as 1,000 people in a state
- 14:48can actually be a mathematically representative sample.
- 14:52Now you also see what are what are called margins of error.
- 14:57Not really explained what these are, but these are based on a statistical concept that we call
- 15:04a confidence interval.
- 15:06So we can't be a 100% confident about the samples results being completely representative
- 15:12of the population, generally representative, closely representative, but not fully
- 15:17representative, because you don't know what the overall population believes.
- 15:22You would have to poll the entire population to do that.
- 15:25So what's really being said is, let's say that a candidate has 45% support and the margin of
- 15:34error is 3%, so
- 15:37the idea is the percentage of the population supporting candidate a is going to be between
- 15:4442 to 48%, but there's still
- 15:47an outside chance that maybe 49% of people support that candidate or 50% of people support
- 15:54that candidate.
- 15:55So what's really being said, depending on what the confidence interval is, and don't need to
- 16:00go too far into that, but typically you see confidence intervals of 95 99 percent cent
- 16:06statistically.
- 16:07So you're what you're really saying is we are 95% confident that the percentage of the
- 16:13population that supports candidate a or ballot measure, whichever, is between 42% 48%.
- 16:22So it's most likely going to
- 16:25be in that range, but there's still an outside chance that is not the case.
- 16:30And, again, things can change over time, so you have to account for that.
- 16:35And then oftentimes, a sample is treated as monolithic by media headlines and regular
- 16:41people.
- 16:41We also wanna know what specific demographic subgroups are.
- 16:45For example, what do black voters think?
- 16:47What do female voters think?
- 16:48What do evangelical white males think?
- 16:51Specifically, we
- 16:52need to know those kinds of things as well because certain demographics are more
- 16:58predisposed to vote for another candidate.
- 17:01For example, black voters overwhelmingly vote democrat.
- 17:06If support for a democratic candidate by the black population is lower than we would expect,
- 17:12that democrat might have some trouble.
- 17:15Depends on other factors like how many Democrats and how many Republicans, or
- 17:21Democratic and Republican voters can vote in that race,
- 17:23but if support from a demographic subgroup is lower
- 17:29or higher than expected, that can change how much or what exactly that poll might be telling
- 17:36us.
- 17:38The media rarely looks beyond the surface level stats, and if they do, there isn't much depth.
- 17:43Though, in fairness, lots of media consumers aren't that interested in that really nitty
- 17:48gritty detail, and media only has so much time to cover these things.
- 17:52They can't just devote 5, 10 minutes to a poll, necessarily.
- 17:58They gotta move on to the next topic of discussion.
- 18:02Now that said, not every poll is worthwhile.
- 18:05Media often treats every poll as valuable, or biased media outlets who are clearly one side
- 18:11or the other often like to cherry pick polls that are favorable to their candidates.
- 18:15Not always.
- 18:15Sometimes they will point out polls that are not favorable to their candidate,
- 18:21and there's even been instances of purely fake polls getting traction.
- 18:27It's not very common, but you
- 18:28do see it every once in a while.
- 18:30Back in 2016, I believe, Debbie Stabenow, believe there was a poll regarding her election
- 18:37that said Kid Rock had more support theoretical support.
- 18:42He didn't actually declare his candidacy for that race, but he theoretically had more
- 18:47support than Debbie Stabenow did.
- 18:50Problem is that poll was completely made up.
- 18:54They didn't actually
- 18:57they I forgot
- 18:58the exact details, but the group that did the poll was not real, and they didn't actually
- 19:04poll anybody.
- 19:06But they gained traction.
- 19:08So you gotta be gotta look out for that, and you gotta make sure that there are reputable
- 19:12resources or, polling firms that have been around for a while and have been fairly
- 19:18accurate in the past.
- 19:19It's not very often that you see false polls, but every once in a while, it does come up.
- 19:25And there also have been some struggles for polling in recent years.
- 19:29For example, increasing nonresponse rates.
- 19:32So nonresponse is you try to make contact with a respondent.
- 19:36The way we we get samples is we randomly select people based on we might have their phone
- 19:43numbers or, you know, or whatever way we have to contact them.
- 19:49You randomly select them.
- 19:50You don't look at 1 and go, we're gonna pick this one.
- 19:53That bias is the sample.
- 19:55Being random makes it so that it's very likely to be representative sample.
- 20:02So non responses are where you try to make contact with a respondent, but they, for
- 20:06whatever reason, don't connect.
- 20:08Maybe they just don't wanna talk to you and they tell you that.
- 20:10Maybe you don't even get ahold of them at all.
- 20:13And that can bias the sample because the pool of potential respondents gets smaller, and it
- 20:17really messes with the random nature of the sample.
- 20:20So and those nonrespondents, a lot
- 20:24of them are gonna be voting in that election, so you're not necessarily getting their
- 20:29opinion, and that might slant the results.
- 20:32Also, polls are getting increasingly more expensive.
- 20:35More questions is a higher cost, especially when you have higher quality polls.
- 20:41Higher quality also costs more.
- 20:43You yeah.
- 20:44You ideally, you would wanna be able to hit people's brains about all these different
- 20:51aspects
- 20:51of the election, but you only have so much time, so much money, so much manpower, people
- 20:57actually working for that firm.
- 20:59Respondents don't wanna sit there all day and answer questions, So more respondents is also
- 21:06more costly, especially if you keep getting non responses because it's just more time.
- 21:11So polling has also had that lately, but high quality polling is incredibly important for
- 21:19these campaigns.
- 21:20Now what are ideal measures for a poll?
- 21:22So for population, because of the way statistics work, at least 1,000 respondents is
- 21:28a pretty good sized poll.
- 21:29Obviously, you'd like more if if you can.
- 21:32Lower is not necessarily a bad thing, though.
- 21:34It just depends on what the election is, but 1,000 respondents tends to be a pretty good
- 21:40number of people.
- 21:41Just if you actually dive into the statistical procedures, it turns out 1,000 is actually a
- 21:48pretty good sample, even for a really big population.
- 21:53Margin of error, you're usually looking for 3 to 5%.
- 21:57If it's higher than that, it's probably a little I
- 22:00don't know if we wanna really
- 22:02not necessarily that those aren't useful, but you want that 3 to 5% range.
- 22:07And then you want each demographic subgroup being represented proportionally if you wanna
- 22:12know multiple demographics opinions.
- 22:15Now if there is
- 22:16a little bit of difference in so, for example,
- 22:20if you have a sample where the overall population of Hispanics in whatever state it
- 22:26is, is 25%, and that state is having a governor's race.
- 22:31But you have a poll that is 20% Hispanic respondents, so there's a 5% difference or a 5
- 22:39percentage point difference.
- 22:40So pollsters will do what's called weighting, where they will add a little bit more weight to
- 22:47the Hispanics respondents because they are not fully it's not a proportional representation in
- 22:54the sample.
- 22:55And how you weight things, it depends on the polling firm.
- 22:59They all have their own ways of doing it.
- 23:01There's not a specific correct way to do it, but sometimes you'll see that happen so that
- 23:08they can try to more accurately gauge what a demographic subgroup's opinions are about a
- 23:14candidate or a ballot measure.
- 23:16And then the last point about polling, we also often see what's called a likely voter, and
- 23:22that's a gauge of the likelihood that a poll respondent will vote.
- 23:26Just as a respondent act answers the poll does not mean that they're actually going to vote.
- 23:32Now problem is there's no objective definition of who a likely voter is.
- 23:38Pollsters have their own formulas for how to determine it, but it's basically based on
- 23:43several data points that include, but are not necessarily limited to, that voter's previous
- 23:47voting history.
- 23:48If they vote in every single election that they can possibly vote for, they vote for
- 23:53presidential elections, they vote in midterms, there's odd year elections that they also
- 23:58participate in.
- 23:59They're probably gonna show up for this election.
- 24:01Now if it's somebody who only sporadically votes, they voted in this the previous
- 24:07president election, and they could have voted in the one before that but didn't, then you
- 24:12might have a little bit of, I don't know if they're gonna necessarily be a likely voter.
- 24:18Voters' responses on the poll question, if they seem like they're particularly energized and
- 24:23wanting to turn out for an election, you might consider them a likelier voter, and then you
- 24:29might put more weight on a demographic that is expected to see an increase in participation.
- 24:35For example, you might see potentially more black voters coming out because Kamala Harris
- 24:40is a black woman too, so you might consider them to be more likely voters depending on how
- 24:47they respond to a question.
- 24:48So it's it's not an exact science, but it is a way that we try to gauge who's actually going
- 24:57to come out to vote, because if they're not gonna come out to vote, you're probably not
- 25:01that concerned with what their opinion is because they can't help you.
- 25:05They can't hurt you, but they can't help you either.
- 25:08Next factor, the numbers and percentages of Democrats, Republicans, and other voters in a
- 25:15state or a district.
- 25:17So the partisan affiliation is the strongest indicator of how an individual votes.
- 25:22Regardless of what happens between
- 25:24Kamala Harris and Donald Trump, regardless of anything that they say or do, most Democrats
- 25:30Okay.
- 25:31Tomala Harris, he's the Democrat.
- 25:33Most Republicans are gonna vote for Donald Trump because he's a Republican.
- 25:39Now for some states, it's easier to track the party affiliation than others because 30 plus
- 25:44states track the affiliation of
- 25:46the voters.
- 25:47The other states just register their voters
- 25:50and then and then that's it.
- 25:52Now the parties also have their ways of, if they make a lot of contact with certain voters,
- 25:58they may have them in their own databases as, okay.
- 26:01This person's a Republican.
- 26:03Okay.
- 26:03This person's a Democrat.
- 26:05It's not necessarily an a an official designation, but parties sometimes in their
- 26:11databases of voter contacts, depending on their previous inter interactions with these voters
- 26:18or certain tells that they give, like if
- 26:20they have a bunch of bumper stickers that suggest that they're a
- 26:24Democrat or Republican, you might have them as that too.
- 26:28But for actual departments of elections, for governments, they might they're not necessarily
- 26:34tracked by the party affiliation of these voters.
- 26:37Now if the partisan lean for an electorate is heavy enough in a given contest, it's extremely
- 26:42difficult for the party with the disadvantage in numbers to win, an exception being Maryland
- 26:48is otherwise a very Democratic state, but Larry Hogan, a Republican, won the governorship in
- 26:542014 and 2018, currently in Kansas, which is an otherwise Republican state.
- 27:00Laura Kelly, Democrat, has won the last two governor's races there, She's, at least at the
- 27:07time of this recording, at the time of the 2024 election, she is the incumbent governor
- 27:12in Kansas.
- 27:14So examples of states with a heavy partisan
- 27:17lean that do track their voters' party affiliation, On the Republican side, where
- 27:22there's a heavy plurality or majority, Idaho is a majority of Republicans.
- 27:28Wyoming is a huge majority of Republicans.
- 27:30There's over 80% that are registered with the Republican party.
- 27:34The Dakotas, Utah, Oklahoma, if they're not a majority, they're
- 27:40at least a plurality where they're the largest subgroup.
- 27:43And then on the Democratic side, place states that have heavy plurality or majority, Maryland
- 27:50is a majority.
- 27:51California is a sizable population.
- 27:54It's not quite a majority, but it is a sizable population, and now independents actually make
- 28:00up more than Republicans in California.
- 28:04Delaware, New York.
- 28:06And then there's still a few states that have a plurality of independence.
- 28:11And then, of course, you have third parties, so there's relatively few people that register
- 28:15under a third party label, like libertarian, constitution, green, and these third parties
- 28:21are not recognized as a ballot qualified party in every single state.
- 28:26So you're ballot qualified, meaning you're not gonna see a candidate that is registered with
- 28:32that party, in a state where that state does not recognize them as a ballot qualified party.
- 28:40Libertarian is the most common third party, but you'll see some states where libertarians are
- 28:45not recognized as ballot qualified.
- 28:47They might have them as a you you you would need a certain amount of support to qualify as
- 28:53a ballot qualified party, and that the the criteria for how you do that, it differs from
- 29:00state to state.
- 29:00And then sometimes, there are different designations for independence depending on the
- 29:04state.
- 29:05Alaska has 2 categories known as nonpartisan and undeclared.
- 29:10Nonpartisan is what it sounds like where you don't register with a party, but you
- 29:15specifically say, I'm not registering with
- 29:17a party.
- 29:18Undeclared is, I'm registering the vote.
- 29:21I'm not even giving any indication
- 29:24of what way, Aileen.
- 29:25I'm not even gonna say I'm nonpartisan.
- 29:26I'm just a voter is basically what it means.
- 29:30And these independents are not necessarily swing voters either.
- 29:33Some independents actually vote more like, quote, unquote, strong partisans, strong just
- 29:38meaning that they often vote with the Republican party or they often vote with the
- 29:43Democratic party, but because of their independent nature, sometimes it's hard to
- 29:49determine how exactly they lead.
- 29:53So it might appear that they're swing voters.
- 29:56The the group as a whole might be a swing bloc, but individual independents aren't necessarily
- 30:02swing voters, if that makes sense.
- 30:05Next factor, concurrent elections and candidates.
- 30:08So elections happening at the same time can help each other out, or they can drag each
- 30:12other down.
- 30:13So there might be a prominent ballot measure bringing in more voters.
- 30:16Here in my state of Florida, the abortion ballot measurement is expected to bring in more
- 30:22voters and expected to favor the Democratic Party
- 30:27in that regard.
- 30:28We'll see
- 30:28if that actually happens.
- 30:30Candidate a's controversy might cause issues for candidate b in the same party even if
- 30:36they're in a different race, for example, Donald Trump trying to win North Carolina.
- 30:40Right now, governor can't gubernatorial candidate Mark Robinson for Republicans is in a
- 30:47scandal.
- 30:48For for some people, Mark Robinson is seen as a drag on Donald Trump, whereas in other races,
- 30:55Donald Trump might be seen as a drag on a gubernatorial candidate.
- 30:59Or on the flip side, Kamala Harris might be seen as a boon to a Democratic candidate, while
- 31:06for another Democratic candidate, it might drag them down.
- 31:08And then for some voters, they either reward or punish candidates of the same party up and down
- 31:12the ballot simply because they are the same party.
- 31:16So we talk about straight ticket voting versus split ticket voting.
- 31:19Straight ticket voting means you specifically or you vote for one party on every partisan
- 31:26election that is on that ballot.
- 31:28You don't unlike in a split tickets voting situation where one contest, you vote for a
- 31:35Republican, next contest, you might vote for a Democrat, next one, you might vote Democrat,
- 31:38then another one, you might vote
- 31:39Republican.
- 31:41For a
- 31:41lot of voters, they will just vote that same party pretty well.
- 31:45And there are 7 states that allow a straight ticket voting option where you can just simply
- 31:49make one mark, and it applies to every candidate in that party where you just fill in
- 31:55that mark, and then, okay.
- 31:57We recognize that you're just gonna vote for every Republican candidate on the ballot, but
- 32:02you're not gonna fill in bubble for each individual one.
- 32:05Same thing with if they do a straight ticket voting option for Democrats.
- 32:09And then the final factor we'll go over is campaign strategies, current events, and party
- 32:14spending.
- 32:14So we're gonna kinda lob those together.
- 32:17So if the parties are pouring more money into a race, it's likely that the race is more
- 32:21competitive.
- 32:22I'm talking about if the Democratic Party itself pours more money into a race or the
- 32:27Republican Party itself pours more money into the race.
- 32:30Of course, the individual candidates' campaigns are gonna try to spend money, but the parties
- 32:36themselves or specific political groups might get involved too if they think that there's a
- 32:42reasonable chance that they're gonna win.
- 32:44Now it could be an incorrect estimation, and that party and their candidate still gets blown
- 32:50out, but that that might not necessarily be the case.
- 32:53Same thing with independent expenditure committees, the super packs that can bring in
- 32:59unlimited contributions and can spend unlimited amounts of money, but they can only do that if
- 33:05they don't coordinate with the campaigns that they're supporting.
- 33:09So if they're getting more heavily involved, it's likely that that contest is going to be
- 33:16more competitive.
- 33:17And then campaign strategies encompass a wide range of aspects, like what methods they use to
- 33:21try to appeal to voters.
- 33:22So if they go use more TV versus radio versus mailers, that gets really nitty gritty, and
- 33:29it's it's hard to make predictions on things like that, but that might that might come into
- 33:36play for a predictor or for a forecaster that's trying to make a prediction about that
- 33:41election, what geographical areas they go to.
- 33:43So if a presidential candidate doesn't go enough to a toss-up state and a forecaster
- 33:50thinks they should go to that, they might ding them a bit on the prediction and say, okay.
- 33:56This candidate is less likely to win this state, for example.
- 34:00And then what a campaign thinks the biggest issues are versus what the forecaster thinks
- 34:08the biggest issues are
- 34:10might not necessarily align.
- 34:12So if a candidate doesn't put enough emphasis on abortion and the forecaster thinks this is
- 34:20going to be a critical issue for that state, then they may say that candidate is less likely
- 34:28to win because they're not putting enough focus on abortion.
- 34:32And you can substitute abortion with any issue that you think is particularly important,
- 34:37unemployment, for example, or just economy in general.
- 34:40And then, obviously, current events play a key role.
- 34:42They can make or break a candidate's chances in a close race.
- 34:47It's probably not gonna impact a less competitive race enough to matter, but for a
- 34:53competitive race, it could mean everything.
- 34:57And that is going to do it for today's topic of how forecasters make their predictions about
- 35:02elections.
- 35:03Let's end it off with a random question, just like we always do.
- 35:08Today's random question is, if you could have an exotic animal as a
- 35:14pet that normally you'd legally not be allowed to have as a pet, what would you choose and
- 35:19why?
- 35:20For me, I
- 35:21don't know if you ever seen them, but puffins up in Canada, they're adorable.
- 35:25I believe they also make cute noises.
- 35:28And, I mean, just just look up what a puffin looks like.
- 35:31It's it's it's adorable.
- 35:32So I'd love to have one of those,
- 35:35as a pet.
- 35:36Obviously, not really normally allowed.
- 35:39If
- 35:40if there were no issues with it and it wasn't a hindrance to the muffin, I'd love to have one
- 35:46as a pet.
- 35:48So that is going to
- 35:49do it for today's episode.
- 35:50As always, thank you so much for tuning in, listening, and or watching, and you can follow
- 35:56me on Facebook at Paul Raider, comma, author and speaker.
- 36:00LinkedIn, Medium, I'm also on, my website, paulrader.org, obviously, this podcast, and my
- 36:08book, Why Independents Rarely Win Elections and How They Could Become More Competitive, also on
- 36:13Amazon, available for purchase in ebook and paperback format, and I believe that is it.
- 36:19So gonna leave you
- 36:20off with that.
- 36:21Once again, thank you so much for tuning in, and until next time, take care.